B

IF

a

ARRA

KBCNPTT

C

1d

Renewable and Sustainable Energy Reviews 16 (2012) 3621– 3647

Contents lists available at SciVerse ScienceDirect

Renewable and Sustainable Energy Reviews

j ourna l h o mepage: www.elsev ier .com/ locate / rser

iodiesel production from non-edible plant oils

vana B. Bankovic-Ilic, Olivera S. Stamenkovic, Vlada B. Veljkovic ∗

aculty of Technology, University of Nis, 16000 Leskovac, Bulevar oslobodjenja 124, Serbia

r t i c l e i n f o

rticle history:eceived 28 January 2012eceived in revised form 28 February 2012ccepted 2 March 2012

a b s t r a c t

Because of biodegradability and nontoxicity biodiesel has become more attractive as alternative fuel.Biodiesel is produced mainly from vegetable oils by transesterification of triacylglycerols. From economicand social reasons, edible oils should be replaced by lower-cost and reliable feedstocks for biodieselproduction such as non-edible plant oils. This paper reviews various methods for biodiesel productionfrom common non-edible oils employing alcoholysis reactions. The aim of this paper is to present the

eywords:iodieselatalysison-edible oilsrocess optimizationransesterification

possibilities of the use of non-edible oils into biodiesel production, to consider the various methods fortreatment of non-edible oils and to emphasize the influence of the operating and reaction conditionson the process rate and the ester yield. The special attention is paid to the possibilities of optimization,kinetics and improvement of biodiesel production from non-edible oils.

© 2012 Elsevier Ltd. All rights reserved.

wo-step processes

ontents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36222. Feedstocks and product characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3622

2.1. Types of feedstocks for biodiesel production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36222.2. The most-used non-edible oils for biodiesel production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36232.3. Comparison between alkyl esters of edible and non-edible oils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3624

3. Transesterification methods for modification of non-edible oils into biodiesel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36243.1. Homogeneously catalyzed transesterification processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3625

3.1.1. One-step processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36253.1.2. Two-step (acid/base) processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3628

3.2. Heterogeneously catalyzed transesterification processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36303.2.1. One-step processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36303.2.2. Two-step (acid/base) processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3633

3.3. Enzyme-catalyzed transesterification processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36333.4. Supercritical transesterification processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3635

4. Optimization of non-edible oils transesterification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36384.1. The statistical optimization of non-edible oils transesterification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36384.2. The kinetics of non-edible oils transesterification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36394.3. The possibilities for improvement of non-edible oils transesterification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3642

4.3.1. In situ processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36424.3.2. The use of ultrasonic irradiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36424.3.3. The use of microwave heating. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36424.3.4. The effect of co-solvent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3644

5. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Acknowledgement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

∗ Corresponding author.E-mail address: veljkovicvb@yahoo.com (V.B. Veljkovic).

364-0321/$ – see front matter © 2012 Elsevier Ltd. All rights reserved.oi:10.1016/j.rser.2012.03.002

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3644. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3644

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3644

3 tainab

1

htrbgTdamacasapararp

pciBoobthp

pfugtonaaojrflolFecmFiei

(rfwuma

622 I.B. Bankovic-Ilic et al. / Renewable and Sus

. Introduction

During the past decades worldwide petroleum consumptionas permanently increased due to the growth of human popula-ion and industrialization, which has caused depleting fossil fueleserves and increasing petroleum price. On the other hand, com-ustion of fossil fuels contributes most to emissions of greenhouseases, which lead to atmospheric pollution and global warming.he transport section is almost utterly dependent on petroleum-erived fuels. The increase of the number of transport vehicles,s predicted [1], could affect the stability of environment and cli-ate on the whole planet. Therefore, there is a great awareness in

diesel fuel substitution at the present all over the world with alean, renewable fuel such as biodiesel, which has a lot of technicaldvantages over fossil fuels such as lower overall exhaust emis-ion and toxicity, biodegradability, derivation from a renewablend domestic feedstock, negligible sulfur content, superior flashoint and higher combustion efficiency. The biodiesel could be useds pure fuel or as blend with petrodiesel, which is stable in allatios. Biodiesel production is expected to encourage employmentnd economic development in rural areas, to develop long termeplacement of fossil fuels, to reduce national dependency on theetroleum import and to increase the security of energy supply [2].

Although biodiesel has a lot of advantages in related toetrodiesel, the high price of its production is the main barrier to itsommercial use. Zhang et al. [3] showed that the price of biodiesels about 0.5 US$/L, compared to 0.35 US$/L for petroleum diesel.iodiesel price depends mainly on the cost of feedstocks, the pricef which makes 70–95% of the total biodiesel cost [4–7]. The usef cheap non-edible oils can be a way to improve the economy ofiodiesel production and its commercial production at the indus-ry scale. Because of different climate conditions, various countriesave been looking for various types of non-edible vegetable oils forossible use in biodiesel production.

Currently, edible oils are the main resources for world biodieselroduction (more than 95%) [6]. However, there are many reasonsor not using edible oils as feedstocks in biodiesel production. These of edible oil in biodiesel production has an influence on thelobal imbalance to the market demand and the food supply byheir high prices, the reduction of food sources and the growthf commercial plant capacities. Thus, focus should be shifted toon-edible resources, which are not used in the human nutritionnd could grow in the barren lands. Oils from these resources are,s a rule, unsuitable for human consumption due to the presencef toxic compounds. For example, the main toxic compounds inatropha plants are protein curcin and purgative agents, proteinicin in castor plant, glucoside cerberin in fruits of sea mango andavonoids pongamiin and karajiin in karanja oil [6]. There are somether reasons for biodiesel production from non-edible oils, besidesow cost and impossibility of their use for the human consumption.irstly, there are many oil plants producing large amounts of non-dible oils in nature all over the world. Then, non-edible oil plantsan be easily cultivated in lands unsuitable for human crops withuch lower costs than those of the edible oil crops’ plantation [6].

inally, the growing of these plants reduces concentrations of CO2n the atmosphere [8]. However, as a serious drawback, most non-dible oils contain a high content of free fatty acids (FFAs), whichncreases the biodiesel production cost [7].

Chemically, biodiesel is the mixture of fatty acid alkyl estersFAAEs), most often methyl or ethyl esters (FAMEs and FAEEs,espectively) obtained by the alcoholysis of triacylglycerols (TAGs)rom vegetable oil and animal fats, or more precisely alcoholysis,

ith an alcohol (methanol or ethanol). In the reversible and consec-tive alcoholysis reaction, one mole of acylglycerols reacts with oneole of alcohol and one mole of ester is formed at every step in the

bsence or presence of a catalyst. Alcoholysis of vegetable oil can

le Energy Reviews 16 (2012) 3621– 3647

be chemically or enzyme catalyzed. Chemical catalysts (base andacid) of alcoholysis can be homogeneous or heterogeneous, whilethe enzymes used as a catalyst are lipases. Non-catalytic alcoholy-sis reactions occur at high temperatures and pressures and still donot have any practical application.

A number of reviewing papers in the last few years indicatethe importance of non-edible oils for the future biodiesel produc-tion. These papers have touched many relevant topics for the use ofnon-edible oils in biodiesel production. Balat and Balat [1], Demir-bas [9], Kralova and Sjöblom [10] and Yusuf et al. [11] considerthe different techniques to solve the problems encountered withthe high fuel viscosity. Gui et al. [6], Demirbas [9], Karmakar et al.[8], Kumar and Sharma [12] and Balat [4] describe the physico-chemical characteristics of selected non-edible oil plants and theiroils as well as esters produced from them. The engine performancesof both non-edible oils and biodiesels produced from them havebeen reviewed in several reports [5,10,13–16]. Some review papersdiscuss different approaches of reducing the content of FFAs in var-ious feedstocks via catalytic and non-catalytic transesterification[4,5,7,9,10,16–21]. The cost and environmental impact of biodieselproduction processes is discussed by a few researchers [12,22].A special attention has been paid to biodiesel production fromjatropha oil in India, Malaysia and Indonesia [18,19,23–25]. How-ever, the methods of biodiesel production, the impact of reactionconditions on the overall process rate and the ester yield as wellas the optimization, kinetics and improvement of biodiesel pro-duction from non-edible oils have not yet attracted the attentionthey deserve. Murugesan et al. [26] analyze the effects of opera-tion variables on the transesterification reaction of non-edible oils,while Leung et al. [7], Vyas et al. [21] and Shahid and Jamal [20]focus on specific processes for biodiesel production from differ-ent feedstocks such as Biox co-solvent, supercritical, microwave,ultrasound and in situ processes.

This paper is a review on various methods for biodiesel produc-tion from common non-edible oils employing alcoholysis reactions.The aim of this paper is to present the possibilities of the use ofnon-edible oils in biodiesel production, to consider the differentmethods for treatment of non-edible oils as alternative feedstocksand to emphasize the influence of some operating and reaction con-ditions on the process rate and the ester yield. The special attentionis paid to the possibilities of optimization, kinetics and improve-ment of biodiesel production from non-edible oils.

2. Feedstocks and product characteristics

2.1. Types of feedstocks for biodiesel production

Feedstocks for biodiesel production can be traditionally cat-egorized into three main groups [8]: vegetable oils (edible andnon-edible), animal fats and waste cooking oils (used oily mate-rials). Additionally, algal oils have been emerging in recent yearsas the fourth category of growing interest because of their high oilcontent and rapid biomass production [4,5,16,21,27].

Different kinds of vegetable edible oils, depending upon the cli-mate and soil conditions, are being used as the main conventionallyfeedstocks for biodiesel production such as rapeseed oil in Canada,sunflower oil in Europe, soybean oil in US, palm oil in SoutheastAsia, coconut oil in Philippines, etc. However, the rapidly grow-ing world population and extensive human consumption of edibleoils can cause some significant problems, for instance, starvation indeveloping countries. Therefore, non-edible plant oils become very

promising alternative feedstocks for biodiesel production becauseof large demand for edible oils as food, the higher prices of edibleoils than that of fossil fuels and the lower cost of non-edible oil plantcultivation. However, there will always be a competition between

I.B. Bankovic-Ilic et al. / Renewable and Sustainable Energy Reviews 16 (2012) 3621– 3647 3623

Table 1Oil content in seeds and kernels of some non-edible plants.

Botanical name Local name Oil content, % References

Seed, wt.% Kernel, wt.%

Jatropha curcas Jatropha, ratanjyot 20–60 40–60 [16]Pongamia pinnata Karanja, pungam 25–50 30–50 [16]Azadirachta indica Neem 20–30 25–45 [16]Madhuca indica Mahua 35–50 50 [12,16]Schleichera triguga Kusum 10.65 [52]Ricinus communis Castor 45–50 [16]Linum usitatissimum Linseed 35–45 [16]Cerbera odollam (Cerbera manghas) Sea mango 54 6.4 [12,16]Gossypium sp. Cotton 17–25 [53]Nicotiana tabaccum Tobacco 36–41 17 [54]Argemone mexicana Mexican prickly poppy 22–36 [12,55]Hevea brasiliensis Rubber tree 40–60 40–50 [12,16]Melia azedarach Persian lilac 10 [12]Simmondsia chinensis Jojoba 45–55 [12,16]Thevetia peruviana Yellow oleander 8.41 67 [12]Moringa oleifera Moringa 33–41 2.9 [16]Thlaspi arvense Field pennycress 20–36 [16]Euphorbia lathyris 48 [56]Sapium sebiferum 12–29 [56]

ewi

ok[straia

gfaaBit

btgfwtpWrcrp

fvrfdut

Pistacia chinensis

Datura stramonium

dible and non-edible oily plants for land available. When decidinghich type of oil crops should be grown, in addition to profits, their

mpact on the environment should also be taken into account [28].There are a large number of oil plants that produce non-edible

ils. From a list of 75 plant species containing oil in their seeds orernels more than 30%, 26 species were reported by Azam et al.29] as potential sources of the oil that can be employed for theynthesis of FAMEs suitable for the use as biodiesel. The advan-ages of non-edible oils as a diesel fuel are liquid nature portability,eady availability, renewability, lower sulfur and aromatic contentsnd biodegradability, while their disadvantages are higher viscos-ty, lower volatility and higher percentage of carbon residue as wells reactivity of unsaturated hydrocarbon chains [16].

In relation to vegetable oils, animal fats such as tallow, whiterease or lard, chicken fat and yellow grease are often pricedavorably for conversion into biodiesel, providing an economicdvantage [30]. However, there is a limited amount of animal fatsvailable, so they will never be able to meet the world’s fuel needs.esides, it is the fact that animal fats can create a big problem dur-

ng the biodiesel production since they became solid wax at roomemperature [31].

Waste cooking oils could be a good choice as feedstocks foriodiesel production because they are either priceless or cheaperhan virgin vegetable oils (2–3 times) [32]. Their amount can bereat in each country and is dependent on the use of vegetable oilsrom which they are generated. The transesterification process ofaste cooking oils into alkyl esters reduces the molecular weight

o one-third, the viscosity by about one-seventh and the flashoint slightly, whereas the pour point is considerably increased [1].aste cooking oils can be contaminated by many types of impu-

ities from the cooking process (polymers, FFAs, etc.) and theironversion to biodiesel is complicated. Many review papers areelated to the use of waste cooking oils as feedstocks for biodieselroduction [33–36].

The amounts of oily crops, both edible and non-edible, animalats and waste cooking oils are limited, so it is unlikely to pro-ide worldwide biodiesel production demand. The search for otherenewable sources is needed to provide the required amount of oily

eedstocks. In recent years a high interest has arisen towards pro-ucing biodiesel from microalgae. The advantages of microalgaesing for biodiesel production are: much higher biomass produc-ivities than land plants, some species can accumulate up to 20–50%

30 [57]10.3–23.2 [58]

TAG, no agricultural land is required to grow the biomass and theyrequired only sunlight and a few simple and cheap nutrients [37].

2.2. The most-used non-edible oils for biodiesel production

Various oils extracted from seeds or kernels of non-ediblecrops are potential feedstocks for biodiesel production. The impor-tant non-edible oil plants are jatropha [38,39], karanja [38,40],tobacco [41–43], mahua [44,45], neem [46], rubber [47], sea mango[48], castor [49], cotton [50,51], etc. Of these feedstocks, jatropha,karanja, mahua and castor oils are the most often used in biodieselsynthesis.

In many countries, edible oils are not produced in enoughamounts to meet the requirements for human use and mustbe imported. Hence, the price of biodiesel produced from edi-ble oils is much higher than that of petrodiesel. An interestingcase of biodiesel production from edible vegetable oils is India,where about 46% of the needed amounts for the domestic require-ments are imported [23]. Therefore, non-edible oils from jatropha,karanja, neem, mahua and other plants are the only possibility forbiodiesel production. Table 1 presents a summary on the oily plantsand the oil contents of their seeds or kernels based on a wide lit-erature survey. Only several, most important and most frequentlyused non-edible oils were selected for further discussion.

Jatropha plant is one of the most promising potential oil sourcesfor biodiesel production in South-East Asia, Central and SouthAmerica, India and Africa. Today, it is the major feedstock for pro-duction of biodiesel in developing countries like India, where theannual production is about 15,000 t [23]. It can grow almost any-where, on waste, sandy and saline soils, under different climaticconditions as well as under low or high rainfall and frost. Its culti-vation is easy, without intensive care and minimal effort. Its healthylife cycle of 30–50 years eliminates the yearly replantation. Jatrophaoil content varies depending on the types of species, but it is about40–60% in the seeds and 46–58% in the kernels [12]. Jatropha hascomparable properties to diesel, such as calorific value and cetanenumber [59]. It has a great potential as an alternative fuel since itdoes not require any modification of the engine [23]. The serious

problem with jatropha oil is its toxicity to people and animals [1].

Karanja is a nitrogen-fixing tree producing seeds with a signif-icant oil content. It is a leguminous, medium size hardy tree withfast growing, which is native to India, the United States, Indonesia,

3624 I.B. Bankovic-Ilic et al. / Renewable and Sustainable Energy Reviews 16 (2012) 3621– 3647

Table 2Physico-chemical properties of methyl esters from different edible and non-edible oils.a

Vegetable oil Density, kg/m3 Kinematic viscosityat 40 ◦C, mm2/s

Cetane number Cloud point, ◦C Flash point, ◦C Oxidation stability at110 ◦C, h

Edible oilsLinseed 892 3.75 – −3.8 – 0.2Rapeseed 882 4.44 54.4 −3.3 – 7.6Sunflower 880 4.44 49.0 3.4 183 0.9Soybean 884 4.04 45.0 1.0 178 2.1Peanut 883 4.90 54.0 5.0 176 2.0Palm 876 5.70 62.0 13.0 164 4.0Rice bran 885 4.96 – 0.3 0.4Coconut 807 2.73 – 0.0 110 35.5Olive – 4.50 57.0 – 178 3.3

Non-edible oilsCastor 899 15.25 – −13.4 – 1.1Jatropha 880 4.80 52.31 2.7 135 2.3Karanja 4.80 55.84 – 150Mahua 850 3.98 56.61 – 208 –Neem 884 5.21 57.83 14.4 – 7.1Tobacco 888 4.23 51.6 – 165.4 0.8

1

Asoue

rcyihc

pwantcd

Sah[

Mcwa

EcR[a

o[SnvA

Standard EN14214 860–900 3.5–5.0 >5

a Based on the data taken from [6–8,10,16,29,54].

ustralia, Philippines and Malaysia. Karanja oil has many toxic sub-tances that do not allow its use as a cooking oil. Annual productionf karanja oil in India is 55,000 t [23], of which only 6% is currentlysed [40]. It can be cultivated to improve the soil quality, and thexhausted land can be reused for the agricultural purpose [6].

Castor plant is easily grown as a weed and has similar ecologicalequirements as jatropha. It is native to India, China, Brazil, someountries of ex-USSR and Thailand. India produces about 0.73 Mtearly, which is 60% of the world castor production [12]. Castor oils completely soluble in alcohols and has a viscosity up to 7-timesigher than that of other vegetable oils [60]. The high viscosity ofrude castor oil is a problem for its direct use as a fuel.

Mahua and neem trees, medium to large trees found in mostarts of India and Burma, are a significant source of oils. Almost thehole tree of neem is usable for various purposes in the medicine

nd as pesticides and organic fertilizer. The annual productions ofeem and mahua oil in India are 100,000 t and 180,000 t, respec-ively [23]. Kernels of mahua contain up to 50% of oil [16], whichontains about 20% FFAs making the procedure for biodiesel pro-uction from this oil very required [44].

The major regions of kusum tree production are located in India,ri Lanka, Timor and Java. Bitter in taste and toxic, kusum oil hasbout 5–11% FFA. The pure oil cannot be used as a fuel because of itsigh viscosity. The production of kusum oil in India is 25,000 t/year23].

Sea mango tree, well-known as a “suicide tree”, grows in India,adagascar, Malaysia, Sri Lanka and Cambodia. Having a high oil

ontent (54%), its seeds are a worth source for biodiesel production,hich has been used so far only in a transesterification process with

specific solid catalyst [48].Greece and Turkey are two the main producers of cotton in the

uropean Union and world. Cottonseeds are a source of one of theheapest vegetable oil having extreme viscosity and high density.efined cottonseed oil production in Turkey in 1997 was 20,546 t61]. However, crop production is inconsistent according to harvestrea and climatic conditions.

Several non-edible oils have been isolated from various plants allver the world such as Moringa oleifera [62–64], Argemone mexicana55], Thevetia peruviana [65], Thlaspi arvense [66], Euphorbia lathyris,

apium sebiferum [56], Pistacia chinensis bge [57] or Datura stramo-ium [58]. Beside many medicinal uses and significant nutritionalalue, M. oleifera oil has great potential for biodiesel production.fter extraction of vitamin A and other high valuable constituents,

0.3 max >101 >6

the oil can be converted to biodiesel without any waste [63]. T.arvense, known as field pennycress or stinkweed, is native to NorthAmerica and Eurasia. It is highly adapted to a wide variety of cli-matic conditions, tolerant of fallow lands and requires minimalagricultural inputs (fertilizer, pesticides, water). Meals from defat-ted seeds cannot be used as an animal feed because of high contentof glucosinolate. Because of high oil content (20–36%), the field pen-nycress seeds are an acceptable feedstock for biodiesel production[66]. E. lathyris and S. sebiferum produce also significant amounts ofoil in their seeds. Primarily, the both plants are native to China andadapted to alkaline, saline, droughty and acidic soils [56].

2.3. Comparison between alkyl esters of edible and non-edible oils

Some of the important physico-chemical properties of methylesters produced from different edible and non-edible resources areshown in Table 2. A high density of esters from some edible (linseed)and non-edible (castor) oils is attributed to the presence of fattyacids having more than two double bonds [8]. However, havinga higher content of saturated fatty acids, methyl esters producedfrom neem oil have a much higher cloud point than those producedfrom linseed oil. The high flash point of mahua biodiesel is due tothe presence of esters with chains having more than 12 carbons.Coconut biodiesel has the smallest flash point. Further, biodieselproduced from some non-edible oils (jatropha, neem) has a smallcontent of linoleic and linolenic acids, so its oxidation stability ishigher than that of biodiesel from linseed oil. Due to the presenceof saturated fatty acids, esters produced from coconut oil have thehighest oxidation stability [8].

Some advantages of biodiesel fuel produced from various non-edible oil feedstocks, compared to diesel fuel, in compressionignition engines are shown in Table 3. The main advantages of thebiodiesels are lower particulate matter, carbon monoxide, hydro-carbons, nitrogen oxides and smoke emission, higher compressionratio associated with higher injection pressure and higher diffusioncombustion.

3. Transesterification methods for modification ofnon-edible oils into biodiesel

Vegetable oils can be converted into biofuel using fourways: blending, micro-emulsions, pyrolysis and transesterification[67,68]. The reversible transesterification reactions are the most

I.B. Bankovic-Ilic et al. / Renewable and Sustainab

Table 3The advantages of biodiesel fuel produced from various non-edible oils.

Oil plant Fuel advantages

Jatropha Lower PM emission, higher compression ratio associatedwith higher injection pressure, higher diffusioncombustion

Karanja Lower NOx emissionMahua Lower CO, HC, NOx emissions, reduced smoke numberRubber Lower smoke emissionNeem Lower CO, NOx and smoke emissions

Pc

ctmeam

aotTtTttbp

3

ooaat

3

tovctboorahitTsyp

chterc

M – particulate matter, NOx – nitrogen oxides, CO – carbon monoxide, HC – hydroarbon.

ommon method of converting TAGs from oils into biodiesel andhe most promising solution of the high viscosity oil problem. The

ain factors affecting transesterification reaction and producedsters yield are: the molar ratio of alcohol:oil, type of alcohol, typend amount of catalyst, reaction temperature, pressure and time,ixing intensity as well as the contents of FFAs and water in oils.The transesterification reaction can be non-catalyzed or cat-

lyzed by an acid, a base or an enzyme. Depending on the solubilityf the chemical catalyst in the reaction mixture, transesterifica-ion reaction can be homogeneously or heterogeneously catalyzed.hese reactions can be accomplished as one-step (base or acid) orwo-step (acid/base) processes, depending on the content of FFA.he latter process is recommended if a feedstock contains morehan 1% of FFA [44], although some authors have recently suggestedhe content of 5% of FFA as the limit for process choice [69]. It shoulde mentioned that there is the method of transesterification that iserformed under supercritical conditions.

.1. Homogeneously catalyzed transesterification processes

Homogeneously catalyzed alcoholysis of non-edible oils viane- and two-step processes is the subject of the most researchesn the FAAE synthesis and is the most frequently industriallypplied biodiesel production. The feedstocks characteristics suchs its FFA content and fatty acid composition influence dominantlyhe selection of the type of the biodiesel production process.

.1.1. One-step processesThe choice between base or acid catalysts mostly depends on

he FFA content (or acid value) in the oily feedstock. The acid valuef a feedstock is influenced by the oil source, the type of culti-ation and the oil storage mechanism. Non-edible oils, generally,ontain significant amounts of FFA that limit the use of high effec-ive base catalysts. Wide ranges of acid value for non-edible oils cane found in the literature. For example, the acid value of jatrophail varies from 0.92 mg KOH/g [70] to 28 mg KOH/g [7]. FFAs fromily feedstocks in the presence of base catalysts form soaps, whicheduce FAME yield, cause catalyst loss and complicate phase sep-ration. Base catalysts are preferable in the case of vegetable oilsaving a lower FFA content. Acid catalysts have low susceptibil-

ty to the presence of FFA in the oily feedstock due to their abilityo simultaneously catalyze reactions of both FFA esterification andAG alcoholysis. On the other hand, the acid catalyzed alcoholy-is reaction is slow and long time is required to reach a high FAAEield. Therefore, acid catalysts have been rarely applied in one-steprocesses.

Table 4 summarizes the catalyst type and the optimal reactiononditions for some homogeneously catalyzed (base and acid) alco-olysis of non-edible oils. Independently, of the type of catalyst,

he alcoholysis reactions occur with the use of methanol or rarelythanol at temperatures below the boiling point of alcohol. Sulfu-ic acid and alkali hydroxides (NaOH and KOH) are the most-usedatalysts in transesterification processes. Almost in all reported

le Energy Reviews 16 (2012) 3621– 3647 3625

studies the ester yield higher than 90% is achieved, independentlyof the type of used catalyst, alcohol and plant oil.

Beside a lot of advantages, the use of homogeneous catalystshas many drawbacks. Operating problems are related to the useof alkali hydroxides as a catalyst because they are hazardous [17].Additionally, in order to meet the specified product quality, theprocess involves a number of washing and purification steps pro-ducing a large amount of wastewater, which is environmentallyunfavorable and requires appropriate treatment. The high amountof water used in washing and consequent treatment of the resultinglarge effluent increases the overall process cost. For these reasons,homogeneously catalyzed alcoholysis could be considered as a tra-ditional method for biodiesel synthesis, and alternative methodshave been developed.

Base catalysis. Base catalyzed transesterification reaction isoften used for biodiesel production from non-edible oils (Table 4).Base catalysts are highly catalytically active and low cost whilebiodiesel of high quality is obtained in a short reaction time [17].The main disadvantages of base catalysts are the impossibility ofconverting FFA to alkyl esters and soap formation in the presenceof FFA, which reduces the biodiesel yield and prevents glycerolseparation [18].

The high biodiesel yield (higher than 90%) was achieved in themost of the studies, independently of the type of feedstock andalcohol. In the case of castor oil ethanolysis using sodium ethox-ide as a catalyst and high alcohol:oil molar ratio (16:1), the highestester yield of 99% was achieved [86]. Also, in the methanolysis reac-tion of jatropha and tobacco oils (the molar ratio 5.6:1 and 10:1,respectively) a high ester yield of 98% was obtained in 1.5 h [71]and 5 min [84], respectively. At the other side, very low yields ofbiodiesel were reported for jatropha [72,77] and sea mango [48]oils having a high FFA content.

So far, most researchers have used methanol and rarely ethanolfor biodiesel production because the methanolysis is faster thanethanolysis and FAME yields are higher than those of FAEE yields.In the case of biodiesel production from castor oil, the maximumFAME yield of 90% is achieved in 1 h at 60 ◦C while the maximumFAEE yield of only 80% is obtained in 5 h at 80 ◦C [60]. However,a mixture of methanol and ethanol was used to produce biodieselfrom jatropha oil in the presence of ultrasound [79]. In this way,the advantage of better oil solubility in ethanol than in methanolwas used. Also, the mixed esters are a better solubility additive thanmethyl esters [79].

The alcohol:oil molar ratio usually used in the alcoholysis reac-tion of non-edible oils is 6:1 [42,60,76], although some researcherssuggest a much higher molar ratio such as 10:1 [46,81,84] oreven 113:6 [88]. The researchers generally agree that the increaseof the initial alcohol:oil molar ratio increases the ester yield[62,72,81,83,84,89]. However, da Silva et al. [86] showed thatthe ester yield did not increase when the alcohol:oil molar ratioincreased above the molar ratio of 16:1. In the case of castor oilethanolysis, a high yield is obtained at higher ethanol:oil molarratios and lower catalyst concentrations or at lower ethanol:oilmolar ratios and higher catalyst concentrations, independently ofthe reaction temperature [60,86].

Most used base catalysts in transesterification processes areNaOH and KOH. The initial catalyst concentration is a very impor-tant factor influencing the TAG conversion degree. The optimalamount of the base catalyst is about 1% (based on oil weight),although some researchers have reported both slightly lower cata-lyst concentrations such as 0.5% [84], 0.7% [46] or 0.8% NaOH [73],and higher catalyst concentrations such as 3.3% NaOH [72] or 6%

NaOH [60]. When the NaOH concentration was 3.3%, the maximumFAME yield was only 55%, which was explained by the coexistenceof FFA in the oil [72]. The FFA content in crude jatropha oil wasreduced from 3.1% to 0.25% in the presence of NaOH under the

3626I.B.

Bankovic-Ilic et

al. /

Renew

able and

Sustainable Energy

Review

s 16 (2012) 3621– 3647

Table 4A review of the homogeneously one-step transesterification processes of different non-edible feedstocks.

Feedstock (oil) Type, volume of reactor, cm3/Typeof agitator, agitation intensity, rpm

Type of alcohol Alcohol:oilmolar ratio,mol/mol

Catalyst/loading,wt.% to the oil

Temperature, ◦C Optimal reaction conditions Reference

Reaction conditions Yield(conversion),%/Time, min

Jatropha Flask, 1000/– Methanol 5.6:1 NaOH/0.5–1.5 30–60 60 ◦C; 1% NaOH 98/60 [71]Pilot plant, 250a/– Methanol 5.6:1 NaOH/0.5–1.5 30–60 60 ◦C; 1% NaOH 96/90Glass tube, 15/Magnetic, 400 Methanol 0.1–0.7:1b NaOH/0.5–3.0 65 0.7:1b; 3.3% NaOH 55/120 [72]Batch reactor, –/–, 300 Methanol 9:1 NaOH/0.8 45 (96.3)/30 [73]

Ethanol 9:1 NaOH/0.8 45 (93.14)/30– Methanol 12:1 NaOH/1 65 89.7c/7 [74]Batch reactor, –/–, 5000 Methanol KOH/– 23 68/120 [75]Batch reactor, 224/Mechanical, 900 Methanol 6:1 KOH/1 40–60 50 ◦C 97.1/120 [76]Flask, 500/Mechanical, 600 Methanol 6:1 NaOH/1 60 47.2d/60 [77]

0.4:1j H2SO4/1 60 92.8d/240Double-necked flask,100/Magnetic, –

Methanol 3:1–10:1 KOH/1 28 and 45 45 ◦C; 6:1 (95)/180 [78]KOH/1 28 and 45 45 ◦C; 6:1 (99)e/180

Flask, 50/– Mixture ofmethanol andethanol3:3 mol/mol

6:1 KOH/0.75 98d/7 [79]

JatrophaKaranja

Pilot plant tank, –/Mechanical, –

MethanolMethanol

6:16:1

KOH/1KOH/1

6565

91/18089/180

[80]

Karanja – Methanol 3:1 and 10:1 KOH/1 45 and 60 10:1; 60 ◦C (92)/90 [81]Methanol 10:1 KOH/1 60 (95)e/90

– Methanol 27:1 KOH/1e 60 – [82]

Neem – Methanol 10:1 NaOH/0.7 60–75 88–94/6.5–8 h [46]Ethanol 10:1

Flask, –/–, 450–500 Methanol 4:1; 6:1; 8:1 KOH/1–3 55, 60, 65 6:1; 60 ◦C; 2% KOH 83.4/60 [83]

Tobacco Flask, 1000/– Methanol 6:1 NaOH/<1.5 55 (86)/90 [42]Flask, 1000/–, 600 Methanol 4:1–10:1 KOH/0.5–1.5 40–60 10:1; 40 and 50 ◦C; 1% KOH 98/5 [84]

Methanol 4:1–10:1 NaOH/0.5–1.5 40–60 50 ◦C; 10:1; 0.5% NaOH 98/5

Castor Batch reactor, 250/Magnetic, – Methanol 6:1 NaOH, KOH, NaOCH3,KOCH3/0.2f

60 KOCH3 85/60 [60]

Ethanol 6:1 80 NaOCH3 80/300Flask, 250/Magnetic, 600 Ethanol 12:1–20:1 NaOC2H5/0.5–1.5 30–80 30 ◦C; 16:1; 1% 93.1/30 [85]Flask, 250/Magnetic, 400(laboratory)

Ethanol 6:1–39:1 NaOH andC2H5ONa/0.5–1.5

30–80 30 ◦C; 16:1; 1% C2H5ONa (99)e/30 [86]

Batch reactor, 1000/Mechanical,400 (large scale)

19:1 NaOH/1 20–80 30 ◦C 89.8/10

Mechanical shaker 9.65:1–12.35:1 KOH/1.33–2.17 30 11:1; 1.75% KOH 80/90

Soybean and castor (25:75,w/w)

Batch reactor, 50/Magnetic, – Methanol 34:6 NaOH/1:6f – 87/1–10 hg [87]

Cottonseed and castor(50:50, w/w)

86/1–10 hg

Castor Batch reactor, 50/Magnetic, – Ethanol 34:6 KOH/1:6f – – 96/120 [88]Soybean and castor (75:25,

w/w)34:6 and 113:6 KOH/1:6f – 113:6 98/120

Sea mango Flask, –/Magnetic, – Methanol 6:1 NaOH/1 64.7 8.3/60 [48]

Cottonseed – Methanol 6:1 KOH/0.5–1.5 50–60 60 ◦C; 1.5% KOH 91.4/30 [50]92.4h/7

Batch reactor, 500/Mechanical, 600and ultrasoundi

Methanol 7:1 NaOH/1–2 60 2% NaOH 95/20, 95i/20 [51]

I.B. Bankovic-Ilic

et al.

/ R

enewable

and Sustainable

Energy R

eviews

16 (2012) 3621– 36473627

Table 4 (Continued)

Feedstock (oil) Type, volume of reactor, cm3/Typeof agitator, agitation intensity, rpm

Type of alcohol Alcohol:oilmolar ratio,mol/mol

Catalyst/loading,wt.% to the oil

Temperature, ◦C Optimal reaction conditions Reference

Reaction conditions Yield(conversion),%/Time, min

Ethanol 7:1 NaOH/1–2 80 1; 2% NaOH 86/60, 90i/60Flask, 1000/ Methanol 3:1–9:1 NaOH/0.5–1.5 60 6:1; 1% NaOH 98.5/60 [89]Magnetic, 400

Mahua Double-necked flask,100/Magnetic, –

Methanol 3:1–10:1 KOH/1 28 and 45 45 ◦C; 6:1 (95)/180 [78]KOH/1 28 and 45 45 ◦C; 6:1 (99)e/180

Flask, 500/– Methanol 4.5:1–9:1 H2SO4/1.0–7.0 65–70 9:1; 6% H2SO4 [90]Ethanol 80–85 92/300Butanol 118–120 95.4/300

Fodder radish Kettle reactor, 1000/Mechanical,500–3200

Ethanol 6:1–14:1 NaOCH2CH3/0.9–1.5 36–60 30 ◦C; 6:1; 1.3% 97.9/70 [91]

Castor Batch reactor, 250/Magnetic, – Methanol 6:1 H2SO4/0.2f 60 85/480 [60]HCl/0.2f 75/240

Ethanol 6:1 H2SO4/0.2f 80 65/240HCl/0.2f 75/480

Moringa oleifera Beker, –/Magnetic, 400 Methanol 1:1–5:1b KOH/0.5–1.5 30–60 60 ◦C; 3:1b; 1% KOH 82/60 [62]

a Capacity, dm3/day.b w/w.c Microwave irradiation.d Ultrasonic power: 210 W.e THF as co-solvent.f Molar ratio.g Average of different reaction times from 1 to 10 h.h 21% of an exit microwave power of 1.2 kW.i Ultrasonic power: 200 W; frequency: 24 kHz.j v/v.

3 tainab

oisdsoatBug

epdrtrtrryob

atmrcttekctt

lttlt[t

tcuit6ea

awtt

3

(wcf

628 I.B. Bankovic-Ilic et al. / Renewable and Sus

ptimum reaction conditions [71]. NaOH neutralizes FFAs andnduces soap formation, so an extra step is needed to remove theodium soap after the reaction. If the NaOH concentration wasecreased below or increased above the optimum, there was noignificant change in the biodiesel yield [71]. Comparing the typef catalyst in the same operating conditions (alcohol type, temper-ture and alcohol:oil molar ratio), Parlak et al. [84] reported thathe same ester yield was obtained using 1% KOH and 0.5% NaOH.eside KOH and NaOH, potassium and sodium methoxides are alsosed as base catalysts. The catalytic efficiency of hydroxide ion wasenerally inferior to that of methoxide ion [60].

Different reaction times for completing the base catalyzed trans-sterification processes of non-edible oils have been reported. Thisrocess requires about 1.5–2 h, independently of the operating con-itions, except much lower (5 min) [84] and much higher (5 h) [60]eaction times in the case of tobacco seed and castor oil, respec-ively. It is known that the conversion rate increases with theeaction time [92]. Chitra et al. [71] used four different reactionimes in order to optimize the transesterification process of jat-opha oil. Keeping methanol amount, catalyst concentrations andeaction temperature constant, the authors showed that the estersield increased with reaction time. Furthermore, although the yieldbtained by Tapanes et al. [73] is slightly lower than that reportedy Chitra et al. [71], the reaction time is three times shorter.

The reaction temperature influences both the reaction ratend the yield of esters [92] and should be maintained belowhe boiling point of alcohol [93,94]. The optimal temperature for

ethanolysis of non-edible oils is about 60–65 ◦C, although someesearchers recommend lower temperature such as 30 ◦C in thease of castor oil [86,95] and 50 ◦C in the case of jatropha [73,76],obacco seed [42,84] and mahua oil [78]. In order to optimizehe reaction temperature, some authors [71,82] have used differ-nt temperatures in the alcoholysis processes of non-edible oils,eeping alcohol amount, catalyst concentrations and reaction timeonstant. Their results clearly indicate that the ester yield propor-ionately increases with the increase in reaction temperature andhe maximum yield of ester is obtained at 60 ◦C.

Chitra et al. [71] performed the biodiesel production on bothaboratory- and large-scale after having optimized the concentra-ion of alcohol, catalyst loading, reaction temperature and reactionime required for the transesterification of jatropha oil. Somewhatower average ester yield was achieved by the large-scale produc-ion than in the laboratory conditions. Using castor oil, da Silva et al.86] performed scale up with 8 times higher oil amount and foundhat the same ester concentration was obtained at both levels.

Acid catalysis. Besides many disadvantages like the slower reac-ion rate, the higher alcohol:oil molar ratio requirement, a loweratalyst activity and a need for higher process temperature, these of acid catalyst in the transesterification reaction has some

mportant advantages [19] such as the tolerance and less sensi-ivity towards the high FFAs presence in the low-cost feedstocks (>%) [21] and the possibility of simultaneously accomplishment ofsterification and transesterification. The most used acid catalystsre sulfuric, phosphoric, hydrochloric, etc.

Biodiesel yield was up to about 90% in several studies related tocid-catalyzed transesterification (Table 4), but a much longer timeas needed, compared to the base catalyzed process, except at high

emperatures [60,77,90]. The type of alcohol, catalyst quantity andhe reaction time are the factors that influence the ester yield [90].

.1.2. Two-step (acid/base) processesTo take advantages of both base and acid catalysts, two-step

acid/base) processes for biodiesel production from non-edible oilsith a high FFA content have been developed. The two-step pro-

ess consists of acid catalyzed FFA esterification (pre-treatment)or reducing the FFAs (below 1%). The base catalyzed TAG

le Energy Reviews 16 (2012) 3621– 3647

alcoholysis is an effective way to achieve a high biodiesel yieldwithin a short reaction time at mild reaction conditions, comparedto one-step process. By using an acid catalyst (mainly sulfuric acid)in the first step of the process, the slow reaction problem is over-come, and the soap formation is eliminated. The only disadvantageof the two-step transesterification process is the higher productioncost as compared to conventional, one-step process. A review oftwo-step homogeneously alcoholysis processes employing differ-ent non-edible oils are presented in Table 5. Important variablesaffecting the acid value in the first and the ester yield in the secondstep are the type of feedstock and alcohol, alcohol:oil molar ratio,catalyst concentration, reaction temperature and reaction time.

Crude and refined, deodorized jatropha oils containing the FFA15% and 1%, respectively, illustrate the significance of the type offeedstock for biodiesel production. The FFA content of the formeroil is far beyond the acceptable limit of base catalyzed transes-terification, and the two-step process is the best way to utilizethis oil. In addition, the refined deodorized jatropha oil has amuch higher price than the crude oil. A comparison betweenthe two different methods of biodiesel production shows thatFAME yield achieved by the two-step acid/base-catalyzed esterifi-cation/transesterification is much higher (90%) than that obtainedby the one-step base-catalyzed transesterification (55%) [72]. Thesimilar results were reported by Deng et al. [77] working with jat-ropha oil with 10.45% of FFAs. In the base catalyzed process, due tosaponification, the ester yield reduced significantly (down to only47.2%), while a much higher ester yield (96.4%) was obtained inthe two step process. To decrease the initial high FFA content incrude mahua oil below 1%, Ghadge and Raheman [44] carried out atwo-stage pretreatment process of esterification catalyzed by sul-furic acid with removal of a methanol–water mixture by settlingafter each step. D’Oca et al. [103] have developed a novel base/acidprocedure for the ethyl esters production from castor oil. The basecatalyzed transesterification is used as the first step followed bythe on pot addition of sulfuric acid, which improves the separa-tion of FAEEs from glycerol because of breaking soap formation andincreases FFAs yield in the esters’ phase. At the end of the secondstep, the acid-catalyzed esterification of FFAs yielded more FAEEs.Afterward, the acid-catalyzed esterification of FFAs from the reac-tion mixture yields more FAEEs, compared to the conventional basecatalyzed transesterification.

Methanol is the mainly used alcohol in both steps of the transes-terification process. An exception is the biodiesel production fromcastor oil [101], where the optimal ethanol:oil molar ratio is muchhigher (40:1 and 20:1 in the first and second stage, respectively)than that in the methanolysis reaction of other non-edible oils. Forthe other non-edible oils, the reaction of methanolysis has beenstudied for different molar ratios, which are varied in the rangeof 6:1 to 18:1 in the both steps, independently of the type of cat-alyst. With increasing the molar ratio of methanol:oil in the firststep, the acid value is sharply reduced at first, then decreases grad-ually and stays constant at the end [40,44,56,58,72,77,98], whilethe ester yield continuously increases [40,96]. The decrease of theacid value is due to the effect of water produced during the esteri-fication of FFA, which prevents further reaction. Therefore, thereis an optimum alcohol:oil molar ratio required to complete theesterification process. In the second step, relatively less amountof methanol is required for reducing the acid value, because someof FFAs have been already esterified during the first step, and lessamount of water is produced during the reaction [44]. The low levelof remained FFA from the first step affects the transesterificationreaction. The esters yield increases to the optimal value and more

or less remains the same with further increasing the methanol:oilmolar ratio [38,43,52,56,58,96,97]. This is attributed to a moredifficult separation of glycerol due to its emulsification with methylesters [97].

I.B. Bankovic-Ilic

et al.

/ R

enewable

and Sustainable

Energy R

eviews

16 (2012) 3621– 36473629

Table 5A review of the two-step (acid/base) catalyzed homogeneously transesterification processes of different non-edible feedstocks (I – first step: acid pretreatment, II – second step: base-catalyzed).

Feedstock (oil) Type, volume of reactor, cm3/Typeof agitator, agitation intensity, rpm

Step Type of alcohol Alcohol:oil molarratio, mol/mol

Catalyst/loading, wt.%to the oil

Temperature, ◦C Optimal reaction conditions Reference

Reaction conditions FFA conversion,%/Yield(conversion),%/Time, min

Jatropha I Methanol 0.2:1–0.4:1a H2SO4/1.3–1.6 60 1.43% H2SO4a (92.8)/88 [39]

II Methanol 0.15:1–0.25:1a KOH/0.55b 60 4:1 (0.16:1a) 99/24Glass tube, 15/Magnetic, 400 I Methanol 0.1:1–0.7:1c H2SO4/1 50 0.6:1c (93)/60 [72]

II Methanol 0.5:1–3.0:1 NaOH/1.4 65 ca. 6.7:1, 0.24:1c 90/120Flask, –/Magnetic, 1000 I Methanol 3:1–12:1 H2SO4/0.3–2 40–100 45 ◦C, 6:1, 0.5% H2SO4, (93)/120 [38]

II Methanol 3:1–12:1 KOH/0.5–2.0 60 60 ◦C, 9:1, 2% KOH 95/120Flask, –/Magnetic, 1000 I Methanol 3:1–9:1 H2SO4/0.25–1.5 40–100 45 ◦C, 6:1, 0.5% H2SO4 (93)/120 [96]

II Methanol 3:1–9:1 KOH/0.5–2.0 40–100 60 ◦C, 9:1, 2% KOH 95/120Batch reactor, 1500/–, 400 I Methanol 2:8–2:3a H2SO4/0.5–3 20–80 65 ◦C, 3:7a, 1% H2SO4, (95)/180 [97]

II Methanol 3:7a NaOH/0.5–3 20–80 50 ◦C, 3:7a, 1% NaOH 90.1/180Flask, 500/Mechanical, 600 I Methanol 0.16:1–0.48–1a H2SO4/1–6a 60 0.4:1a, 4% H2SO4

a (88.5)/60 [77]II Methanol 0.16:1–0.48–1a NaOH/0.8–1.6 60 0.24–1a, 1.4% NaOH 96.4d/30

Flask, 1000/Mechanical, 600 I Methanol 8:1–10:1 H2SO4/0.2–1.0 60 8:1, 0.4% H2SO4 (92)/30 [56]II Methanol 6:1 KOH/0.6–1.2 60 6:1, 1% KOH 86.2/30

Stock tank, 50 I Methanol 0.16:1–0.4:1a H2SO4/1–4a – 0.32:1, 3% H2SO4a (91.4)/– [98]

II Methanol 8:1 NaOH/1.2 – – (99.38)e/10

Jatrophaf Batch reactor, –/Mechanical, – I Methanol H2SO4/– [75]II Methanol KOH/– 23 73/120

Karanja Flask, –/Magnetic, 1000 I Methanol 3:1–9:1 H2SO4/0.25–2 40–80 50 ◦C, 6:1, 1% H2SO4 (94)/45 [96]II Methanol 3:1–9:1 KOH/0.3–1.0 40–80 50 ◦C, 9:1, 0.5% KOH 80/30

– I Methanol 6:1 H2SO4/0.5 65 (91)/– [40]II Methanol 6:1 KOH/1 65 97/–

Flask, –/Teflon stirrer, 300 I Methanol 0.1:1–0.5:1c H2SO4/0.5–5 33.83:1c, 3.73% H2SO4 (81.8)/3.17 [99]II Methanol 0.1:1–0.5:1c KOH/0.5–1.5 9.3:1 (33.4:1c), 1.33% KOH 89.9g/2.5

Jacketed reactor, 1000/Digitalstirrer, 600–650

I Methanol 6:1 H2SO4/1 30 and 60 60 ◦C, 15:1 (95)/– [100]II Methanol 6:1 KOH/1 30 and 60 60 ◦C, 15:1 (97)/60

Mixture of mahua and simaroubaConical flask, 250/–, 230 I Methanol Mahua oil:0.34–0.36a

H2SO4/1.4–1.49a 19.37:1, 14.38% –/59–90 [45]

Simarouba oil:0.27–0.40a

H2SO4/1.4–1.54a –/40–90

II Methanol 5:1 KOH 60 98/30

Kusum Flask, 1000/Mechanical, 1200 I Methanol 6:1–12:1 H2SO4/0.5–1.25a 40–64 50 ◦C, 10:1, 1% H2SO4a (96)/60 [52]

II Methanol 6:1–12:1 KOH/0.5–1.1 40–65 50 ◦C, 8:1, 0.7% KOH 95/60

Tobacco Flask, –/Magnetic, 400 I Methanol 4.5:1–18:1 H2SO4/1–2 60 13:1, 2% H2SO4 (94)/50 [43]II Methanol 6:1 KOH/1 60 91/30

Castor Flask, 100/Magnetic, 300 I Ethanol 40:1 H2SO4/1 60 –/60 [58]II Ethanol 20:1 KOH/1 60 (95.3)/60

Flask, 100/Magnetic, 300 I Ethanol 40:1 H2SO4/1 60 –/60 [101]II Ethanol 20:1 KOH/1 60 (95.6)h/60

Palm/rubber blend (1:1) Flask, –/–, 360 II Methanol 6:1–10:1 KOH/0.5–2 45–65 55 ◦C, 8:1, 2% KOH (96)/3 [102]

Datura stramonium I Methanol 6:1–10:1 H2SO4/0.4–1 60 8:1, 0.6% H2SO4 (89)/30 [58]

3630 I.B. Bankovic-Ilic et al. / Renewable and SustainabTa

ble

5

(Con

tinu

ed)

Feed

stoc

k

(oil

)Ty

pe,

volu

me

of

reac

tor,

cm3/T

ype

of

agit

ator

, agi

tati

on

inte

nsi

ty, r

pm

Step

Typ

e

of

alco

hol

Alc

ohol

:oil

mol

arra

tio,

mol

/mol

Cat

alys

t/lo

adin

g,

wt.

%to

the

oil

Tem

per

atu

re,◦ C

Op

tim

al

reac

tion

con

dit

ion

s

Ref

eren

ce

Rea

ctio

n

con

dit

ion

sFF

A

con

vers

ion

,%

/Yie

ld(c

onve

rsio

n),

%/T

ime,

min

II

Met

han

ol

6:1

KO

H/0

.8–1

.2

60

1%

KO

H

87/3

0

Thla

spi a

rven

seFl

ask,

500/

IM

eth

anol

9.2:

1H

2SO

4/1

a[6

6]M

agn

etic

, 120

0II

Met

han

ol6:

1N

aOC

H2C

H3/0

.5

60

94/9

0

[5p

t]

Mor

inga

olei

fera

II

Met

han

ol

6:1

NaO

CH

3/1

60

–/60

[64]

Euph

orbi

a

lath

yris

Flas

k,

1000

/

I

Met

han

ol

8:1–

10:1

H2SO

4/0

.2–1

.0

60

10:1

, 0.8

%

H2SO

4–/

45

[56]

Mec

han

ical

, 600

IIM

eth

anol

6:1

KO

H/0

.6–1

.2

606:

1,

0.8%

KO

H86

.2/3

0

Sapi

um

sebi

feru

mFl

ask,

1000

/I

Met

han

ol8:

1–10

:1

H2SO

4/0

.2–1

.0

60

8:1,

0.4%

H2SO

4–/

30

[56]

Mec

han

ical

, 600

IIM

eth

anol

6:1

KO

H/0

.6–1

.2

606:

1,

1%

KO

H88

.3/3

0

av/

v.b

w/v

.c

w/w

.d

Ult

raso

nic

pow

er:

210

W.

eM

icro

wav

e

pow

er:

800

W. 2

450

kHz,

con

tin

uou

s

pro

cess

, flow

rate

:

5

(I

step

)

and

3

mL/

min

(II s

tep

).f

Jatr

opha

goss

ypiif

olia

.g

Mic

row

ave

pow

er:

180

W.

hIn

the

pre

sen

ce

of

n-h

exan

e

as

co-s

olve

nt.

le Energy Reviews 16 (2012) 3621– 3647

Independently of the type of non-edible oil, a number ofresearchers have used sulfuric acid as an acid catalyst (the requiredamount varied from 0.4 to 5%) in the first stage and KOH or NaOHas a base catalyst (the required amount varied from 0.5 to 2%) in thesecond stage of the process. The catalyst amount affects the esteryield which firstly, increases and then starts to decline when theacid or base concentration increases above the optimal values (0.5%and 2%, respectively) [38,56,96,98]. A lower amount of acid catalystthan the optimal one in the first step does not reduce the acid valueto the desired limit, whereas a higher amount results in darkeningof the product [52,56,58]. Alternatively, when the transesterifica-tion of pretreated oils is carried out at a base catalyst concentrationhigher than the optimal one, the amount of soap generated duringthe reaction increases, and the ester yield is reduced [52,56].

Experiments with non-edible oils have been performed for var-ious reaction times between 0.4 and 6 h. With the progress of thereaction up to the optimal reaction time, the FAME yield increasesrapidly. In the acid-catalyzed pretreatment, when the reaction timeis longer than the optimal one, the physical appearance and colorof the oils become darker [56].

Most of researchers have been conducted their experiments attemperatures close to the boiling point of alcohol [38,52,96,97]. Themaximum ester yield is obtained in the range of temperature from50 to 60 ◦C. The reaction temperature for processing non-edibleoils should be maintained below 60 ◦C because the saponificationof TAGs by the base catalyst is much faster than the alcoholysisreaction [38,96]. Also, the base-catalyzed process at temperaturesabove 60 ◦C causes excessive methanol loss due to evaporation,reduces the overall ester yield [96] and increases the productioncost [52].

All above-mentioned researches refer to the batch experimentswith mechanical or magnetic agitators. There is only one two-steptransesterification process, which is carried out in the continuousmode [98]. The high TAG conversion (about 99.4%) was obtained ina reactor with microwave irradiation. The reactor was developedto convert jatropha oil to alkyl ester using NaOH as a catalyst andto compare its performance with traditional reactors. It was con-cluded that the two-step continuous microwave process might finda practical application in the biodiesel production from non-edibleoils.

3.2. Heterogeneously catalyzed transesterification processes

3.2.1. One-step processesA biodiesel synthesis by using heterogeneous (solid) catalysts

is environmentally friendly because of a simple product separa-tion and purification, which reduces the waste water amount. Theadditional benefit of the heterogeneous catalyst use is the possi-bility of their easy regeneration and reuse that make the biodieselsynthesis process cost-effective [104]. However, a major disadvan-tage of using heterogeneous catalysts is a low reaction rate causedby diffusion limitations in the three-phase (oil–alcohol–catalyst)reaction mixture [105], as well as the complex catalyst preparationfollowed by a significant contribution to the environmental impactin some cases. Recent researches have been focused towards lowcost and eco-friendly heterogeneous catalysts with a high catalyticactivity [106–109]. Generally, the preparation of this type catalystinvolves washing, drying, crushing/powdering and calcinating athigh temperatures.

Various compounds were investigated as heterogeneous cata-lysts for non-edible vegetable oils alcoholysis. Table 6 summarizesthe catalyst type and the reaction conditions applied in heteroge-

neously catalyzed alcoholysis.

The catalytic activity of a heterogeneous catalyst depends onits nature, specific surface area, pore size and volume and activesite concentration. The catalyst performance could be improved by

I.B. Bankovic-Ilic

et al.

/ R

enewable

and Sustainable

Energy R

eviews

16 (2012) 3621– 36473631

Table 6The review of the catalyst type and reaction conditions applied in heterogeneously catalyzed alcoholysis of non-edible oils.

Feedstock (oil) Type, volume ofreactor, cm3/Typeof agitator,agitation intensity,rpm

Alcohol Alcohol:oil molarratio, mol/mol

Catalyst/loading, wt.%to the oil

Temperature, ◦C Optimal reaction conditions Reference

Reaction conditions Yield (con-version),%/Time, h

Jatropha Batch reactor Parr4843,50/Mechanical, 300

Methanol 4:1–14:1 MontmorilloniteKSF/1–20

80–180 12:1; 4.8%; 160 ◦C 68/6 [110]

Amberlyst 15/1–20 59/6Round-bottomflask, 50/Magnetic

Methanol9:1, 16:1, 20:1 K/NaY zeolite (Kloadings of 4, 8 and12 wt.%)/4

65 16:1; 12 wt.% K on NaY 73/3 [111]

Round-bottomflask, 150/Magnetic

Methanol 5:1–25:1 CaMgO/1–10 65 15:1; 4% (83)/6 [112]

CaZnO/1–10 (81)/6Flask, 500/300–700 Methanol 3:1–15:1 KNO3/Al2O3/1–8 70 12:1; 6%; 600 rpm (84)/6 [113]Round-bottomflask, 50/

Methanol 55:1 Mg–Zr mixed oxidewith Mg/Zr weightratios of 1:1, 2:1 and3:1/10

30 and 65 Mg/Zr = 2:1; 65 ◦C ≈(100)/0.75 [114]

Batch reactor Parr4842,300/Mechanical

Methanol 9:1–18:1 Mg0.7Zn1.3Al2/3O3/1.5–10.5 150–190 11:1; 8.68%; 182 ◦C 94/6 [115]

Flask, 100 Methanol 6:1–15:1 CaO/0.5–2 50–70 9:1; 1.5%; 70 ◦C (93)/2.5 [116]Round-bottomflask/Mechanical,300

Methanol 6:1 CaO, Li–CaO;La2O3–ZnO;La2O3/Al2O3;La0.1Ca0.9MnO3;CaO + Fe2(SO4)3;Li–CaO+ Fe2(SO4)3/5

60 CaO + Fe2(SO4)3;Li–CaO+ Fe2(SO4)3

≈100/3 [117]

Round-bottomflask, 10

Methanol 1:2a Mg/La mixed oxideswith different Mg/Laweight ratios/5

Ambient and65

Mg/La = 3:1; 65 ◦C ≈100/0.5 [118]

Jatropha (pretreated) Erlenmeyer flask,25; Ultrasound24 kHz, 200 W,Amplitude 30–70%,Pulse mode0.3–0.9 s/s

Methanol 5:1–15:1 Na/SiO2/1–5 – 50% amplitude; 0.7 s/s;9:1; 3%

98.5/0.25 [119]

Jatropha Flat bottom SS316stainless steelreactor,100/Magnetic, 170

Methanol 15:1 Dodecatungestophosphoric(DTPA) supported onK-10 clay/5

170 20 wt.% DTPA/Clay (93)/8(93)/8

[120]

Karanja Round-bottomflask

Methanol 3:1–15:1 Li/CaO with Li ionamount from 1 to3 wt.%/1–10

35–85 Li/CaO with Li ionamount of 1.75 wt.%;12:1; 5%; 65 ◦C

>99/2>99/1

[121]

Karanja – Methanol 10:1 Montmorillonite K-10;H�–Zeolite; ZnO/0.1

120 ZnO (83)/24 [81]

Parr reactor,500/600

Methanol 6:1–15:1 Li/CaO; Na/CaO;K/CaO/0.5–10

30–65 Li/CaO; 12:1; 2%; 65 ◦C 94.9–90.3b/8 [122]

Cerbera odollam Batch reactor Methanol 10:1 MontmorilloniteKSF/4;

155 SO42−/ZrO2/6; 8:1;

180 ◦C84/3 [48]

8:1 SO42−/ZrO2/6 180

3632I.B.

Bankovic-Ilic et

al. /

Renew

able and

Sustainable Energy

Review

s 16 (2012) 3621– 3647

Table 6 (Continued)

Feedstock (oil) Type, volume ofreactor, cm3/Typeof agitator,agitation intensity,rpm

Alcohol Alcohol:oil molarratio, mol/mol

Catalyst/loading, wt.%to the oil

Temperature, ◦C Optimal reaction conditions Reference

Reaction conditions Yield (con-version),%/Time, h

Castor Batch reactor,100/Magnetic

Methanol 6:1–35:1 Zn5(OH)8(NO3)2·2H2O/5 40–60 29:1; 60 ◦C 20/3 [123]

Stainless steelreactor

Methanol 6:1 TiO2/SO42− with molar

ratio TiO2/H2SO4 of 5,10 and 20/5c

120 TiO2/SO42− with molar

ratio TiO2/H2SO4 = 525/1 [124]

Flask,250/Mechanical;Flask, 250,Microwave (40 and220 W)

Methanol 6:1 SiO2/(30–50)% H2SO4;Al2O3/50% KOH/10

Ambient and60

MW irradiation 40 W;Al2O3/50% KOH

(95)/0.08 [125]

Ethanol 6:1 SiO2/(30 and 50)%H2SO4/10

MW irradiation 40 W;Al2O3/50% H2SO4

(>95)/0.33

Round-bottomquartz flask, 250,Microwave(2.45 GHz,220 W)/Magnetic,600

Methanol 3:1–18:1 (40–60)% H2SO4/C/1–7 55–70 12:1; 55% H2SO4/C; 5%;65 ◦C

94/1 [126]

Round-bottomquartz flask

[127]

Conventionalheating,250/Mechanical,600

Methanol 18:1 NaHSO4·H2O, AlCl3,Na2CO3/7.5

65 Na2CO3 79/4

Microwave(2.45 GHz,200–600 W)/Magnetic,600

Methanol 6:1–30:1 NaHSO4·H2O, AlCl3,Na2CO3/1.5–9.5

55–85 Na2CO3; 18:1; 7.5%;65 ◦C; 220 W

90/2

Cottonseed Autoclave, 250 Methanol 3:1–18:1 TiO2-SO42−;

ZrO2-SO42−/2

200–230 TiO2-SO42−; 12:1;

230 ◦C>90/8 [128]

Cottonseed Batch reactor,250/Mechanical

Methanol 6:1–18:1 KF/�-Al2O3 withKF-Al2O3 mass ratio of20.21 to50.36%/1–5

50–68 KF-Al2O3 mass ratio of50.36%; 12:1; 4%; 65 ◦C

>90/3 [129]

Used cottonseed Round-bottomflask

Methanol 3:1–15:1 Li/CaO, Na/CaO, K/CaOwith Li ion amountfrom 0.5 to 5 wt.%/1–8

35–75 Li/CaO with Li ionamount of 1.5 wt.%;12:1; 5%; 65 ◦C

>99/0.75 [130]

Acidified cottonseed Autoclave,250/Magnetic

Methanol 3:1–12:1 SO42−/TiO2-SiO2/1–5 180–220 9:1; 3%; 200 ◦C 92/6 [131]

Rubber seedJatrophaCalophyllum L. (pinnai)Karanja

Teflon-lined steelautoclave,100/rotating, 50

Methanol 15:1 K4Zn4[Fe(CN)6]·6H2O·2(tert-BuOH)/3

170 (97)/8(87)/8(84)/8(88)/8

[132]

Yellow oleander Batchreactor/Magnetic

Methanol 20:1d Ash of banana (M.Balbisiana) trunk

32 96/3 [65]

Moringa oleifera Pressurized batchreactor/350–360

Methanol 6:1–24:1 SO42−/SnO2-SiO2/3 60–180 19.5:1; 150 ◦C 84/2.5 [63]

a Mass ratio.b FFA content in oil from 0.48 to 5.75%.c mol.% of catalyst to the oil.d mL/g of oil.

tainab

csiilaautitLc

tiioosanea[croaifa

atmatte

o[afta

sdhoSso(tihfCt8att

I.B. Bankovic-Ilic et al. / Renewable and Sus

atalyst support on the carriers, which provide a higher specificurface area [133] or by applying the appropriate pretreatment toncrease the catalyst acidity or basicity [111,116,134]. The activ-ty of heterogeneous catalysts is attributed to the presence of thearge amount of strong basic [114,118,130], acid [132] or both basicnd acid [115] sites. The catalytic activity of an impregnated cat-lyst depends on the type of the active catalyst as well as on thesed catalyst precursor. Kumar and Ali [130] have reported thathe activity of alkaline ion impregnated CaO depends on the metalon type as well as on the metal salt used in the catalyst prepara-ion. According to their results, Li/CaO obtained by impregnation ofi2CO3 possesses the highest surface area and basic strength, andonsequently, the highest activity in the alcoholysis reaction.

The type of a heterogeneous catalyst for the biodiesel produc-ion from non-edible vegetable oils depends on the FFA contentn the feedstock. The most often used non-edible oil in previousnvestigations of heterogeneously catalyzed alcoholysis is jatrophail (Table 6). As expected, base catalysts are preferable in the casef oil with the lower FFA content [113,116,119]. Due to the acidolid catalysts’ ability to simultaneously catalyze esterification andlcoholysis reaction, they are often used for FAME synthesis fromon-edible oils with the high FFA content. Some researchers alsomployed the modified catalysts, which exhibited dual basic andcidic sites [111,115] or the mixture of acid and base catalysts117], which allow oil conversion to biodiesel in a one-step pro-ess of simultaneous esterification and alcoholysis. The previousesearches of heterogeneously catalyzed alcoholysis of non-edibleils have been aimed to develop a catalyst with a high catalyticctivity. The most-used methods for the catalyst preparation arempregnation and co-precipitation usually by using metal nitrates,ollowed by drying and frequently calcination at a desired temper-ture.

The alkyl esters yield obtained in heterogeneously catalyzedlcoholysis is also influenced by the applied reaction condi-ions. Compared to homogeneously catalyzed reaction, the initial

ethanol:oil ratio, the catalyst loading and the reaction temper-ture are higher and the reaction time is much longer in ordero achieve comparable esters yields. Generally, the optimal reac-ion conditions providing the highest alkyl esters yield must bexperimentally determined.

Generally, the Mg/Zr mixed oxides with a Mg/Zr weight ratiof 2:1 [114], Mg/La mixed oxides with a Mg/La weight ratio of 3:1118], Li/CaO [121,130] and mixture of base (CaO and Li-CaO) andcid (Fe2(SO4)3) catalysts [117] have the highest catalytic activityor non-edible oil methanolysis at mild reaction conditions. Theotal conversion of oily feedstock is achieved at 60–65 ◦C and within

relatively short reaction time (Table 6).The major advantage of heterogeneous catalysts is the pos-

ibility of their reusing, which enables the continuous processevelopment. Most researches of heterogeneously catalyzed alco-olysis deal with the catalyst reusability, and different results arebtained. K4Zn4[Fe(CN)6]·6H2O·2(tert-BuOH) [132], SO4

2−/TiO2-iO2 [131] and TiO2-SO4

2−; ZrO2-SO42− [129] were stable catalysts,

uitable for long-term use, but no catalyst regeneration meth-ds are reported. Aluminum oxide-modified Mg–Zn catalystMg0.7Zn1.3Al2/3O3) can be recycled five times with the drop inhe FAME yield of 3% and 6% during regeneration and reusabil-ty from the first cycle, respectively [115]. An effective and stableeterogeneous catalyst is Mg–Zr mixed oxide, which is used in

our cycles with a marginal decrease of the esters yield [114].aMgO and CaZnO are reused, after regeneration, three and fourimes, respectively, with maintaining the conversion higher than

0% [112]. However, in most cases, the heterogeneous catalystsre recycled up to three times after which a significant decrease inhe FAME yield is observed [113,117,119,120]. Catalyst deactiva-ion can be caused by three main reasons: product and by-product

le Energy Reviews 16 (2012) 3621– 3647 3633

adsorption on the catalyst surface, the active sites leaching into thesolution and the catalyst structure collapse [135]. In order to avoidthe catalyst active sites blocking with organic residues or poison-ing with atmospheric CO2, researchers use different regenerationmethods: washing with methanol and drying [114,115,119], wash-ing with methanol and hexane followed by drying and calcinationat a desired temperature [112], drying and calcination [113] andrecycling the catalyst with glycerol [117].

The laboratory scale continuous process of heterogeneouslycatalyzed alcoholysis of non-edible oils was developed by Sreepras-anth et al. [132]. The process was conducted in a fixed-bed, downflow reactor. The catalyst (K4Zn4[Fe(CN)6]·6H2O·2(tert-BuOH)) wascrushed, sized (10–20 mesh) and placed between ceramic beadsinto the reactor. The reactants (unrefined rubber seed oil and n-octanol at the alcohol:oil molar ratio 15:1) were fed using a syringepump with the total flow rate 2 mL/h. The reaction was carried outat 170 ◦C under atmospheric pressure. The obtained rubber seedconversion was 89.2%, and no loss in the catalytic activity wasnoticed after 52 h [132]. Peng et al. [131] have proposed a con-tinuous biodiesel production process from cheap raw feedstockswith the high FFA content, such as non-edible oils, soapstocks andwaste cooking, oil by solid acid catalysis. The production processwas carried out in a sequence of three reactors with a countercur-rent flow of vaporized methanol. The excess methanol was recycledin a distillation tower, while the oil phase was refined at a biodieselvacuum distillation tower. Using the proposed continuous pro-cess, a biodiesel demonstration production plant having an annualcapacity of 10,000 t was built [131].

3.2.2. Two-step (acid/base) processesSeveral studies deal with the use of heterogeneous catalysts in

the two-step process (Table 7). The used oily feedstocks have highacid values between 10.5 mg KOH/g oil [135] and 24.76 mg KOH/g[139]. Heterogeneous catalysts are used either to catalyze FFAesterification or TGA alcoholysis. According to the best authors’knowledge, a completely heterogeneous two-step process has notbeen developed yet as a step is still homogeneously catalyzed.

The researches have been performed in the last few years tofind a cheaper catalyst or a catalyst with a simple preparationmethod that reduces the biodiesel production cost. For example,CaO prepared from waste chicken eggshells by calcination is usedas a catalyst in the two-step biodiesel production process frommahua [138] and karanja [137] oil, and the FAME yields of 95% areachieved. The simple calcination is used for preparing SO4

2−/TiO2[136] and SO4

2−/ZrO2 [139], which are effective catalysts for FFAesterification reaction in a two-step process providing 9% and 94%FFA conversion, respectively. As in the case of one-step process,catalyst reuse is attractive from the point of the continuous pro-cess development. CaO [138] and hydrotalcite with Mg/Al molarratio 3:1 [135] are reused for 10 runs after washing and calcinationand for 8 runs after washing with ethanol, respectively, withoutany significant loss in the FAME yield. The best reusability perfor-mance is shown in the case of SiO2·HF, which is recycled 30 timeswithout the catalyst regeneration and with the unchanged FAMEyield [134].

3.3. Enzyme-catalyzed transesterification processes

It is well-known that lipases, enzymes that catalyze both trans-esterification and esterification reactions, have a high catalyticactivity in water-poor media. Therefore, lipase-catalyzed transes-terification reaction is carried out in non-aqueous environments

such as solvent-free systems, organic and ionic liquids, gaseousmedia and supercritical fluids. The advantages of this process aresimultaneously catalysis (TAG alcoholysis and FFA esterification),easy recovery of glycerol, the use of feedstock with high FFA

3634I.B.

Bankovic-Ilic et

al. /

Renew

able and

Sustainable Energy

Review

s 16 (2012) 3621– 3647

Table 7The review of the catalyst type and reaction conditions applied in two-step heterogeneously catalyzed alcoholysis of non-edible oils (I – first step: acid pretreatment, II – second step: base-catalyzed).

Feedstock (oil) Type, volume of reactor, cm3/Type ofagitator, agitation intensity, rpm

Step Alcohol Alcohol:oilmolar ratio,mol/mol

Catalyst/loading,wt.% to the oil

Temperature, ◦C Optimal reaction conditions Reference

Reactionconditions

Yield (conversion),%/Time, h

Jatropha Three-neck flask, 250/Mechanical I Methanol 1–14b H2SO4/1 30–70 12 wt.%; 70 ◦C (≈91)a/1 [136]1:1–30:1c SO4

2−/TiO2/1–5 70–140 20:1; 4%; 90 ◦C (≈97)a/2Autoclave/Mechanical, 1500 II Methanol 6:1 KOH/1.3 64 98/0.33Flask, 1000/Magnetic, 400 I Methanol 3:1–12:1c SiO2·HF/2–15 50–60 12:1; 10%; 60 ◦C (96)a/2 [134]

II Methanol 6:1 NaOH/1 60 ≈99.6/2Three-neck flask, 500 in ultrasoundreactor 180–270 W/Mechanical, 600

I Methanol 40e H2SO4/4d 60 (≈88)a/1 [135]II Methanol 3:1–6:1 Hydrotalcite with

Mg/Al molar ratio3/1/0.5–1.25

40–55 210 W; 4:1; 1%;45 ◦C

95/1.5

Karanja Round-bottom flask,1000/Mechanical, 150–900

I Methanol 6:1 H2SO4/1.5, v/v 65 – (≈91)a/1 [137]II Methanol 6:1–12:1 CaO (obtained from

chickeneggshells)/1–3.5

45–75 8:1; 2.5%; 65 ◦C;600 rpm

95/2.5

Mahua Round-bottom flask, 1000/Mechanical,1000

I Methanol 6:1–12:1 H2SO4/0.5–2, v/v 45–65 6:1; 1.5 (v/v); 55 ◦C (91)a/1 [138]II Methanol 5:1–9:1 CaO (obtained from

chickeneggshells)/1.5–3.0

45–70 8:1; 2.5%; 65 ◦C 95/2.5

Neem Round-bottom flask/Mechanical I Methanol 3:1–12:1 SO42−/ZrO2/0.5–1.5 30–90 9:1; 1%; 65 ◦C (≈94)a/2 [139]

II Methanol 6:1 KOH/1 60 95/2

a FFA conversion.b wt.% of methanol to the oil.c methanol:FFA molar ratio.d vol.% of H2SO4 to the oil.e vol.% of methanol to the oil.

tainab

cdottib[[capprfosyl

cbrS[pT[nmaeChsivre[aT

dGbctthte[oraaaa

hnbspi

I.B. Bankovic-Ilic et al. / Renewable and Sus

ontent and a minimal quantity of waste water generated. The mainrawback in the use of the enzyme-based process is the high pricef the enzyme. To obtain a reusable enzyme catalyst for the con-inuous processes, lipases are usually immobilized, which enablesheir recycling, easy recovery and lower costs. The lipase origins the main factor which influences the ester yield. The immo-ilized lipases from different origins, such as Candida antractica49,140–146], Chromobacterium viscosum [146,147], Mucor miehei49], Pseudomonas cepacia [148,149], Pancreatic [150] and Rhizomu-or oryzae [144,151], are most frequently used in non-edible oillcoholysis reactions. Generally, enzyme-catalyzed alcoholysis iserformed with a high lipase amount (about 5–30%). Celite, macro-orous acrylic resin, macroporous anion exchange resin, silica andeticulated polyurethane foam are the most often used carriersor immobilized enzyme. In the case of ethanolysis of jatrophail catalyzed by immobilized lipase of Chromobacterium visco-um [147] and Pseudomonas cepacia on Celite 545 [149] the FAEEield is increased, compared to that achieved by freely suspendedipase.

However, lower linear alcohols, such as methanol and ethanol,onventionally used in the biodiesel process, can deactivate immo-ilized lipases [140]. Ethanol is more facile in the enzyme-catalyzedeaction than methanol due to a lower lipase deactivation degree.ome authors have suggested the use of propan-2-ol, ethyl acetate140] or diethyl carbonate [152] as acyl acceptors in the biodieselroduction from non-edible oils to overcome lipase inactivation.he other solutions include the stepwise addition of methanol57,144,145] and the use of organic solvent such as hydrophobic-hexane [49,153] or hydrophilic t-butanol [142,150]. Approxi-ately, the same ester yields are achieved in the presence and

bsence of an organic solvent [148]. However, the use of solventnables a significant advantage of higher reaction rate. Otherwise,hattopadhyay et al. [150] have tested four different solvents:exane, acetonitrile, toluene and t-butanol. The last one is mostuitable due to good solubility of all reactants and products intot and the highest conversion achieved. Anyway, the use of sol-ent in biodiesel production is not generally recommended foreasons of safety and economy. The repeated use of immobilizednzyme is the important way to decrease the cost of product154]. The review of the reaction conditions of the lipase-catalyzedlcoholysis which provide a significant ester yield is given inable 8.

The optimum temperature for enzymatic transesterification isetermined by the lipase stability and the reaction rate [155].enerally, the enzymatic reaction is performed at temperaturesetween 30 and 50 ◦C, except the lipase-catalyzed alcoholysis ofastor oil, which occurs at 65 ◦C [49]. Water is a very impor-ant factor in the enzyme-catalyzed process, because it influenceshe lipase catalytic activity and stability, contributing to theigher ester yield. Also, the water amount plays another impor-ant role influencing the esters hydrolysis, which decreases thesters yield. According to numerous studies with Candida antractica49,140–143,145], the high ester yield is achieved in the absencef water. Immobilized lipase is operationally stable over twelveepeated cycles of alcoholysis using propan-2-ol and ethyl acetates acyl acceptors [140,141]. There is no significant loss in the lipasectivity in the both cases due to better miscibility of propan-2-olnd ethyl acetate in TAG and their less polarity than linear chainlcohols.

Batch laboratory reactors (types of vial, Erlenmeyer or flask)ave been mainly used for enzymatic biodiesel production from

on-edible oils. Only in a study [142], a one-step continuous packeded reactor has been reported to operate over 500 h without sub-trate conversion loss, which is a very important result for theractical process application. The main disadvantage of this process

s the necessity of solvent recovery.

le Energy Reviews 16 (2012) 3621– 3647 3635

3.4. Supercritical transesterification processes

The transesterification of non-edible oils by alcohols (methanol,ethanol, propanol and butanol) under supercritical conditions hasbeen considered to be the most promising process for biodieselproduction. In this way, the initial reaction lag stage caused by thelow solubility of the alcohol in the oil phase can be overcome, thereaction rate is higher, the reaction time is shorter, separation andpurification of biodiesel are simply, there are no soap formation andno waste generation from catalyst separation and absence of cata-lyst makes the glycerol recovery easier. The main disadvantages ofsupercritical reaction are the easy degradation of produced esters atan extremely high temperature and the high cost of the apparatus.The additional drawbacks are the high pressure and temperatureand the large amount of alcohol used [18]. Co-solvents and sub-critical alcohol could be added into reaction mixture to reduce theoperating costs [21].

The review of the operating conditions applied in supercriticalone- and two-step processes in various batch and continuous reac-tors is given in Table 9. A higher alcohol:oil molar ratio is requiredto push non-catalytic supercritical alcoholysis forward and to com-plete conversion when oil contains a high FFA amount [157]. Theyield of produced esters increased with increasing the molar ratioof alcohol to oil [61,157,158,161,162], perhaps due to the increasedcontact area between alcohol and TAG. A 100% FAME yield wasachieved in the case of lower content of FFA (only 2%) in the feed-stock, compared to those obtained with the oils with higher FFAcontents [158].

Similar results are obtained by the transesterification of non-edible oils with supercritical methanol and ethanol [161–163]. Bothalcohols have similar reactivity at the supercritical temperature andproduce similar contents of the esters. However, the conversiondegree is higher with supercritical methanol than with supercriticalethanol [9,61].

Few researchers have studied the effect of the reaction tem-perature and pressure on supercritical alcoholysis. The increaseof the reaction temperature favorably influences the ester yield,independently of the type of feedstock [9,61,156,158,161–163].The common reaction temperature is higher than the criticaltemperatures of methanol (240 ◦C) and ethanol (244 ◦C). Also,approximately the complete TAG conversion can be achieved ata pressure of 20 MPa [157,161,162]. The increase of the reactionpressure influences the ester yield [157,158]. This effect is proba-bly due to the increase of the density with increasing the pressure,which causes an increase in the solvent power of the supercriticalfluid [157]. In the reaction of crude jatropha oil with micro-NaOHunder the pressure ranging from 4 to 14.5 MPa, only a little differ-ence in the ester yield is observed [156]. Some researchers haveshown that the conversion under supercritical conditions is com-pleted in a very short time and there was no separation problem[157,158] inherent to conventional catalytic processes.

Recently, some researchers have applied a novel two-stepprocess as a promising alternative to the one-step supercriticalmethod. It consists of the hydrolysis of TAGs in subcritical water(on 270 ◦C) and the subsequent supercritical esterification of theproduced and separated fatty acids non-catalytically in supercrit-ical methanol [159,160] or dimethyl carbonate [69]. In this way,the FAME yield was increased, compared to the conventional base-catalyzed method. Also, a more valuable by-product, glyoxal, isproduced instead of glycerol, which is not affected by the high FFAcontent [69]. Also, this product is a non-acidic one and does notto deteriorate a feedstock with a high content of poly-unsaturated

fatty acids. However, the cost of dimethyl carbonate is much higherthan that of methanol or ethanol.

In the transesterification of non-edible oil with supercriticalmethanol in the presence of micro-NaOH as the catalyst, a high yield

3636I.B.

Bankovic-Ilic et

al. /

Renew

able and

Sustainable Energy

Review

s 16 (2012) 3621– 3647

Table 8The review of the reaction conditions of the lipase-catalyzed transesterification processes of different non-edible feedstocks.

Feedstock (oil)Type, volume ofreactor, cm3/Typeof agitator,agitation intensity,rpm

Type of alcohol Alcohol:oilmolar ratio,mol/mol

Lipase source(enzymecommercial name)

Carrier Lipaseloading,wt.% to theoil

Temperature, ◦C Solvent Water, wt.% tothe oil

Optimal reactionconditions

Reference

Reactionconditions

Yield(conversion),%/Time, h

Jatropha Screw capped vessel,50/Reciprocating, 150

Methanol 0.5:1–4:1 Rhizopus oryzae Reticulatedpolyurethane foam

1–10 30 – 0–10b 1:1, 6, 5b 80/60 [144]

0.5:1–4:1 Candida antractica(Novozym 435)

Macroporousacrylic resin

1–10 30 – – 1:1, 2 76/90

Erlenmeyer typeflask, 25/–

Methanol 3:1–5:1 Chromobacteriumviscosum

Silica activatedwith ethanolamine

3–6 – – – 4:1, 5 (84.5)c/0.5 [146]

Screw capped vial,–/Reciprocal, 200

Methanol 1:1–5:1 E. aerogenes Silica activatedwith ethanolamine

20–60 U 30–55 t-butanol 5–20b 4:1, 50, 55 ◦C, 0 94/60 [154]

Screw capped vial,10/–, 180

Diethylcarbonate

1.5:1–7.5:1 Candida antractica(Novozym 435)

5–15 45 – 0–3 3.75:1, 13.7, 0 96.2/13.3 [143]

Screw capped vial,–/–, 200

Ethanol 4:1 Chromobacteriumviscosum

Celite-545 10 40 – 0–2a 0.5a 92/8 [147]

Screw capped vial,–/–, 200

Ethanol 4:1 Pseudomonascepacia

Celite 10 50 – 4–5 5 (98)/8 [149]

JatrophaKaranja

Screw capped tube,–/Magnetically, 150

Propan-2-ol 1:1–5:1 Candida antractica(Novozym 435)

Macroporousacrylic resin

5–30 50 – – 4:1, 10 (92.8)/10(91.7)/10

[140]

JatrophaKaranja

Screw capped tube,–/Magnetically, 150

Ethyl acetate 3:1–13:1 Candida antractica(Novozym 435)

Macroporousacrylic resin

5–30 50 – – 11:1, 10 91.3/1290/12

[141]

Castor Erlenmeyer typeflask, 300/–, 200

Ethanol 3:1–10:1 Candida antractica(Novozym 435)

Macroporousacrylic resin

5–20 35–65 n-hexane 0–10 10:1, 20, 65 ◦C, 0 (81.4)/6 [49]

Mucor miehei(Lipozyme IM)

Macroporous anionexchange resin

n-hexane 3:1, 20, 65 ◦C, 0 (98)/6

Mahua Screw capped vial,5/–, 200

Ethanol 4:1 Pseudomonascepacia

K2SO4

microcrystals10 40 – 2–10 6 (99)/2.5 [148]

Glutaraldehyde (92)/2.5Polypropylene (96)/6

Cottonseed Fixed bed batchreactor

Methanol 1.5:1–6:1 Candida antractica(Novozym 435)

Macroporous resin 1.7 25–50 t-butanol 6:1, 50 ◦C 97/24 [142]

Fixed bedcontinuous reactord

Methanol 6:1 Candida antractica(Novozym 435)

Macroporous resin 50 t-butanol 95/500

Screw cappedvessel,50/Reciprocating,150

Methanol 38:1–1:10 Candida antractica(Novozym 435)

3.5 40 – 4:1 95 [145]

–, –/–, 180 Methanol 1:1–25:1 Pancreatic 0.05–3 25–50 t-butanol 0–30 15:1, 0.5, 37 ◦C, 5 75–80/4 [150]

Rice bran Flask,50/Reciprocating,170

Methanol 4:1–1:3 Candida sp. 99–125 20 30–60 n-hexane 0–60 2:1, 40 ◦C, 20 87.4/12 [153]

Pistaciachinensis bge

Shaking flask, 25/–,160

Methanol 3:1–6:1 Rhizopus oryzae(F-AP15)

Macroporous resinand anion excangeresin

7e 30–45 – 0–80 5:1, 37 ◦C, 20 94/60 [57]

a w/v.b v/v.c Ultrasonic: 200 W, 24 kHz.d Enzyme flow rate: 9.6 mL/h.e IU/g.

I.B. Bankovic-Ilic

et al.

/ R

enewable

and Sustainable

Energy R

eviews

16 (2012) 3621– 36473637

Table 9A review of the optimal operating conditions applied in supercritical transesterification processes of different non-edible feedstocks.

Feedstock (oil) Type, volume of reactor, cm3/Typeof agitator, agitation intensity, rpm

Type of acylacceptor

Acyl acceptor(Water):oilmolar ratio,mol/mol

Catalyst/loading,wt.% to the oil

Temperature, ◦C Pressure, MPa Optimal reaction conditions Reference

Reaction conditions Yield(conversion),%/Time, min

Jatropha Cylindrical autoclave,250/Magnetic, 400

Methanol 18:1–36:1 MicroNaOH/0.2–0.8

240–270 4–14.5 250 ◦C, 7 MPa, 24:1,0.8% NaOH

90.5/28 [156]

Stainless steel vessel, 100/–, 960 Methyl acetate 25:1–59:1 – 300–345 10–30 345 ◦C, 20 MPa, 42:1 (100)/50 [157]Bench-scale reactor,3700/Mechanical, –

Methanol 10:1–43:1 – 239–340 5.7–8.6 320 ◦C, 8.4 MPa, 43:1 100/4 [158]

I step: Water I:217:1 – I: 255–350 I: 25–32 I: 270 ◦C, 27 MPa I: –/25 [69]II step: Dimethylcarbonate

II:14:1 II: 270–350 II: 6–19 II: 300 ◦C, 9 MPa II:97/15a

Tubular reactor, 1000/– I step: Water I: – – I: 250–290 I: 11 I: 270 ◦C I: (92.1)/60 [159]II step: Methanol II:3:1b II: 250–290 II: 11 II: 290 ◦C II:(99)/15a

Tubular reactor, 1000/– I step: Water I: 10:1b – I: 250–310 I: 11 I: 290 ◦C I: (94.8)/60 [160]II step: Methanol II: 3:1–7:1b II: 250–290 II: 11 II: 290 ◦C, 3:1 II:(98.3)/9a

Jatropha Batch reactor, 11/– Methanol 10:1–70:1 – 200–400 20 400 ◦C, 50:1 (95)/30 [161]Ethanol 10:1–70:1 200–400 (94)/20

Karanja Methanol 10:1–70:1 200–400 20 400 ◦C, 50:1 (95)/20Ethanol 10:1–70:1 200–400 (94)/30

Jatropha Batch reactor, Methanol 1:1–6:1 Novozym 435,in SC CO2,10–70

30–60 6.8 45 ◦C, 5:1, 30% (≈48)/600 [161]

7/– Ethanol (55)/600Karanja Methanol (≈48)/600

Ethanol (50)/600

Castor Batch reactor of stainless steel,11/–

Methanol 10:1–70:1 – 200–350 20 350 ◦C, 40:1 (≈100)/30 [162]

Linseed Ethanol 10:1–70:1 200–350 20 350 ◦C, 40:1 (≈100)/40

Castor Tubular rector, 7/– Methanol 2:1–10:1 Novozyme 435in SC CO2, 20

30–70 50 ◦C, 5:1 (≈50)/12 [162]

Ethanol 2:1–10:1 30–70 40 ◦C, 5:1 (≈37)/12

Fodder radish(R.sattivus)

Tubular reactor, 32/– Methanol 39:1 – 298–336 10–18.3 317 ◦C, 18.3 MPa, 39:1 97/27 [163]Ethanol 32:1–52:1 310–325 9–14 319 ◦C, 12.5 MPa, 39:1 97.5/22

Cottonseed Cylindrical autoclave, – Methanol 41:1 – 230,250 250 ◦C (≈98)/8 [61]Ethanol 41:1 230,250 250 ◦C (≈86)/8

Linseed Cylindrical autoclave, – Methanol 41:1 – 230,250 250 ◦C ≈98/8 [9]Ethanol 41:1 230,250 250 ◦C ≈86/8

a Two-step process.b v/v.

3 tainab

oWsto

obiNbwteHs

4

4t

riTwIatcftdreaevc

hoaiotrqi[tIlrywtuqyaruc

638 I.B. Bankovic-Ilic et al. / Renewable and Sus

f biodiesel (90.5%) with the best properties was obtained [156].ith increasing the catalyst loading, the reaction rate increases

harply up to the maximum value and then stays constant. Also,he increase of the reaction temperature has a favorable influencen the FAME yield.

Biodiesel can be also synthesized enzymatically from non-edibleils and methanol or ethanol in the presence of supercritical car-on dioxide [48,162]. The optimum conversion of only about 50%

s obtained under supercritical conditions in the presence of lipaseovozym 435. This is a much lower yield than that obtainedy supercritical methanol/ethanol esterification. Further, workingith jatropha oil, Rathore and Madras [161] have reported that

he conversion of jatropha oil is lower with methanol than withthanol, due to the denaturing effect of methanol on the enzyme.owever, working with castor oil, Varma and Madras [162] have

hown completely contrary results.

. Optimization of non-edible oils transesterification

.1. The statistical optimization of non-edible oilsransesterification

Since the efficiency of FAAE synthesis depends on the appliedeaction conditions, it is of great importance to investigate theirnfluence on the esters yield in order to determine the optimal.he classic way of investigation by varying one process variablehile others are hold constant are time-consuming and expensive.

nstead, the statistical methods, known as design of experiments,re frequently used as a power tool for the process optimization inhe past decade such as the response surface methodology (RSM)ombined with the central composite rotatable design [39,85,95],actorial design 23 [86], Box–Behnken factorial design [98,99] orhe combination of fractional factorial and Doehlert experimentalesign [91], as well as Taguchi technique [102] and generic algo-ithm coupled with artificial neural network [45]. These methodsstimate not only the effects of individual process variables butlso the interactions between them, thus allowing a better knowl-dge of the process, determining the optimal levels of the processariables and developing more economic and more efficient pro-esses.

According to da Silva et al. [85,86], the catalyst loading, the alco-ol:oil molar ratio and their interactions have the greatest impactn the FAEE yield, while the effects of the reaction temperaturend the interaction temperature-catalyst loading were not signif-cant. Cavalcante et al. [95] have reported that the ethanol:castoril molar ratio, the potassium hydroxide catalyst loading, the reac-ion time, as well as the interaction between KOH loading andeaction time, are effective on the esters yield in both linear anduadratic model at the 95% confidence level. The catalyst load-

ng influences the ester yield in a negative manner. Valle et al.91] have carried out a comprehensive study on the influence ofhe process variables on the Raphanus sativus ethyl esters yield.n first stage, using a 2(5−1) fractional factorial design, they estab-ished, contrary to other researchers, that the ethanol:oil molaratio in the range 6:1–14:1 had no significant influence on the estersield, while the most significant variable was agitation intensity,hich affected the reaction negatively. The reaction time, reaction

emperature and catalyst loading were studied in the next stagesing the Doehlert design of experiments. The reaction time anduadratic term of reaction temperature positively influence FAEEield, while the reaction temperature and the quadratic term of cat-

lyst loading have the negative effect on the process. The optimaleaction conditions for achieving the maximum biodiesel yield aresually obtained by using the RSM and the mathematical modeloming out from the optimization study [85,86,91,95].

le Energy Reviews 16 (2012) 3621– 3647

The two-step process of biodiesel synthesis from non-edible oilis rarely statistically optimized. Tiwari et al. [39] optimized themethanol:oil molar ratio, H2SO4 loading and reaction time for FFAcontent reduction in the first step, and methanol:oil molar ratio andreaction time for the methanolysis of pretreated oil in the secondstep. Based on quadratic polynomial model, linear and quadraticterms of all variables and the interaction between methanol:oilmolar ratio and catalyst loading were significant in reducing theacid value. The FAME yield was affected by linear and quadraticterms of methanol:oil molar ratio and reaction time, as well as bytheir interaction. In the optimization of two-step karanja biodieselsynthesis under microwave irradiation, Kamath et al. [99] consid-ered methanol:oil molar ratio, catalyst loading and irradiation time.The statistically significant terms in reducing FFA content were lin-ear and quadratic terms of all variables, the most important factorbeing catalyst loading, which affected the FFA content negatively.The most important factors affecting FAME yield in the secondprocess step are methanol:oil molar ratio, catalyst loading, interac-tion methanol:oil molar ratio-irradiation time as well as quadraticterms of all variables, having a negative effect on the FAME yield.Based on polynomial model and RSM, the optimal reaction condi-tions for both steps were established, and predicted results werevalidated experimentally. Only Liao and Chung [98] optimized themethanol:oil molar ratio, NaOH amount and flow rate for a con-tinuous microwave assisted methanolysis of jatropha oil usingBox–Behnken factorial design. The significant effect on FAME yieldhad all three variables, interactions between the catalyst loadingand the methanol:oil molar ratio and between the catalyst amountand the flow rate, as well as the quadratic term of the catalystamount. Based on the RSM methodology, the maximum FAME yieldwas found to be 99.36% under the following optimal reaction condi-tions: methanol:oil molar ratio of 10.74:1, NaOH amount of 1.26%and flow rate of 1.62 mL/min. Yusup and Khan [102] studied theinfluence of the methanol:oil molar ratio, reaction temperature andKOH loading on the base-catalyzed methanolysis of acid treatedblend of crude palm/crude rubber seed oil, and the optimal com-bination for maximizing FAME yield were established by usingTaguchi technique. For identifying the optimal reaction parame-ters (methanol:oil molar ratio, catalyst loading and the reactiontime) for reducing FFA content in mahua and simarouba oil by anacid-catalyzed pretreatment, Jena et al. [45] developed a genericalgorithm method based on an artificial neural network model.

The statistical methods of optimization have been rarely appliedin heterogeneously catalyzed, lipase-catalyzed and non-catalyzedalcoholysis reactions. A few studies deal with the optimization ofheterogeneously catalyzed methanolysis. An orthogonal test wasdesignated to investigate the influence of the reaction variables onthe jatropha oil methanolysis catalyzed by CaO [116]. The order ofthe reaction variables’ effect on the FAME yield was as follows: cat-alyst loading > reaction time > reaction temperature > methanol:oilmolar ratio. Olutoye and Hameed [115] statistically evaluatedthe FAME production from crude jatropha oil using aluminumoxide modified Mg-Zn catalyst. All investigated reaction variables(methanol:oil molar ratio, catalyst loading and reaction tempera-ture) influenced the FAME yield. The optimization of jatropha oilmethanolysis catalyzed by KSF clay and Amberlyst 15 was per-formed using a 23 full factorial central composite rotatable design[110]. The empirical models for FAME yield as a function of reactiontemperature, methanol:oil molar ratio and catalyst loading weredeveloped. In the case of KSF catalyzed methanolysis, the reac-tion temperature, the methanol:oil molar ratio and their interactionhad a significantly positive effect on the FAME yield, while catalyst

loading had a negative effect. All three variables positively influ-enced FAME yield in by Amberlyst 15 catalyzed methanolysis, butinteraction between temperature and catalyst loading was signifi-cantly negative. The optimum reaction conditions for achieving the

tainab

hu(

etcTcpeqnvscotmadaipycbwrasdFvr

mwrftcbemtstesCmtmshotm[rt6aia

I.B. Bankovic-Ilic et al. / Renewable and Sus

ighest FAME yield were established, and the predicted FAME val-es were in agreement with the experimentally obtained dataTable 6) confirming the model accuracy.

The RSM based on central composite design was used by Sut al. [143], who investigate the influence of five important reac-ion variables for FAME production from jatropha oil using diethylarbonate as acyl acceptor and Novozym 435 lipase as a catalyst.he reaction time, reaction temperature, lipase amount and diethylarbonate:oil molar ratio, as well as the interaction between tem-erature and amount of added water had a significant positiveffect on the FAME yield. The amount of added water and theuadratic terms of all variables except the lipase amount had a sig-ificant but negative effect on the esters production. The optimalalues of process variables were estimated by solving the regres-ion equation. The predicted FAME yield under optimal reactiononditions was 97.7%, which agreed well with the experimentallybtained yield of 96.2%. De Oliveira et al. [49] optimized the reactionemperature, water amount, enzyme concentration and ethanol:oil

olar ratio for castor oil ethanolysis catalyzed by Novozym 435nd Lipozyme IM lipases by applying the Taguchi experimentalesign. In the case of Lipozyme IM catalyzed ethanolysis, the watermount, linear and quadratic terms of ethanol:oil molar ratio andnteractions of enzyme concentration with both the reaction tem-erature and ethanol:oil molar ratio were effective on the FAEEield but all with a negative effect, while only the enzyme con-entration had a positive effect. Regarding the system catalyzedy Novozym 435, only the water amount had a negative effect,hile others linear terms, the quadratic term of ethanol:oil molar

atio and interactions between temperature and water amountnd temperature and ethanol:oil molar ratio were statisticallyignificant in a positive manner. The regression equation pre-icted optimal reaction conditions under which the predictedAEE yield (82% and 99.6%) agreed well with the experimentallyalues of 81.4% and 98% for Novozym 435 and Lipozyme IM,espectively.

Valle et al. [163] optimized non-catalyzed ethanolysis andethanolysis of Raphanus sativus L. oil using the Doehlert designith three and two variables, respectively. Based on the quadratic

egression model, the ethanol:oil molar ratio is the most importantactor influencing the reaction negatively, followed by the reactionime and temperature, both having a positive effect. The statisti-ally significant influence on the FAEE yield had the interactionetween reaction time and temperature and the quadratic term ofthanol:oil molar ratio, both in a negative manner. The studies ofethanolysis showed that the reaction temperature and reaction

ime, as well as the interaction between them had a statisticallyignificant and positive effect on the FAME yield. The optimal reac-ion conditions were established based on the RSM and modelquation. The RSM model was used for the optimization of theubcritical hydrolysis of jatropha oil, obtained by the supercriticalO2 extraction of Jatropha curcas kernels, and the supercriticalethylation of the hydrolyzed oil [159]. In this study, two (reac-

ion temperature and reaction time) and three (reaction time,ethanol:FFA volume ratio and reaction temperature) factors were

tatistically evaluated for oil hydrolysis and methylation of theydrolyzed oil, respectively. The both factors were effective on theil hydrolysis. In the supercritical methylation process, the reac-ion temperature and reaction time had a positive effect while the

ethanol:FFA volume ratio had a negative effect on the FAME yield159]. Chen et al. [160] optimized the subcritical hydrolysis of Jat-opha curcas crude seeds oil and the supercritical methylation ofhe hydrolyzed oil. The highest FFA yield was obtained at 290 ◦C in

0 min under 11 MPa, at water:oil volume ratio of 10:1 and aceticcid amount of 2.5 vol.%. The optimum conditions for the supercrit-cal methylation process (3:1 methanol:FFA volume ratio, 290 ◦Cnd 9 min) resulted in the FFA conversion of 98.3%.

le Energy Reviews 16 (2012) 3621– 3647 3639

Shuit et al. [164] statistically evaluated the production of FAMEin a reactive extraction process of Jatropha curcas L. seeds usingRSM coupled with central composite design. Four different processvariables, namely reaction temperature (30–60 ◦C), methanol:seedratio (5–20 mL/g), H2SO4 loading (5–30 wt.%) and reaction time(1–24 h) were optimized. All variables, as well as the interactionbetween reaction temperature and catalyst loading significantlyaffect the FAME yield in a positive manner, the most important fac-tor being the catalyst loading. The significant effect of H2SO4 wasexplained by its role of catalyst not only in the reactive extractionprocess but also in accelerating the oil extraction due to a bettersolubility of lipid and oil in the acidic solvent. Based on the regres-sion model, the highest FAME yield of 99.8% could be obtainedwith the following process variables: reaction temperature of 60 ◦C,methanol:seed ratio of 10.5 mL/g, 21.8% of H2SO4 and reaction timeof 10 h.

4.2. The kinetics of non-edible oils transesterification

The alcoholysis reaction kinetics is indispensable for the pro-duction process development and the reactor design, operation andscale-up. The fundamental understanding of the alcoholysis reac-tion kinetics is necessary for development of mathematical modelsdescribing the reaction rate and the product yield. The kinetic stud-ies provide parameters that are used for prediction of the reactionprogress under particular reaction conditions, process analysis andcontrol.

In the studies of the homogeneous catalyzed methanolysis twostages are well-recognized: initial heterogeneous stage controlledby the mass transfer rate and a pseudo-homogeneous stage which ischemically controlled. The initial stage is caused by the incompletemiscibility of the nonpolar and polar reactants, and it is observedat lower reaction temperatures and agitation intensities. The masstransfer limitations have been included in modeling the overallprocess kinetics only in a few studies of edible oils methanoly-sis [165,166]. Karmee et al. [82] noticed that the oily feedstock,alcohol and catalyst could qualitatively alert the kinetics of alcohol-ysis reaction. In the case of jatropha oil, the methanolysis reactionwas quite rapid especially at the early reaction stage, compared tomahua and the edible oils [78].

The kinetics of the homogeneously catalyzed alcoholysis reac-tion has been most frequently studied, and different kinetic modelswere used (Table 10). Some complex models are based on thestepwise reversible alcoholysis reaction [82,167] and sometimesinclude side saponification reactions [76]. However, there are sim-ple models that assume the irreversible overall reaction [51,78,86].Bikou et al. [167] studied the effect of water content in ethanolon the cottonseed oil ethanolysis reaction kinetics. The depen-dence of both the reaction rate and equilibrium constants are wellcorrelated. This result allows assessing if it is preferable to use anhy-drous, high-price ethanol or ethanol containing a certain amountof water. For the pongamia oil methanolysis, Karmee et al. [82]observed that the fastest reaction was the forward reaction of thefirst step (methanolysis of TAG to diacylglycerols, DAG), while thesecond step (reaction of DAG to monoacylglycerols MAG) was theslowest step. The most complex kinetics model, which includessaponification reactions of all glycerides and esters, was developedfor the methanolysis of a jatropha oil–waste food oil mixture [76].The comparison of the second-order kinetics based on the step-wise methanolysis reaction, and the new proposed kinetic modelshowed that the latter had smaller deviation from experimentaldata.

Development of simple kinetic models which fitted well exper-imental data is favorable, because they do not require complexcomputation of their parameters. The overall reaction of castor oilethanolysis in the kinetically controlled region was described by

3640I.B.

Bankovic-Ilic et

al. /

Renew

able and

Sustainable Energy

Review

s 16 (2012) 3621– 3647

Table 10A survey on the reaction mechanism and kinetics model of homogeneously catalyzed alcoholysis of non-edible oils.

Feedstock (oil) Alcohol Temperature, ◦C Reaction mechanism Kinetics modelc Reference

Pongamia Methanol 60 Three consecutivereversible reactions

− d[TAG]dt

= k1[TAG][MeOH] − k−1[DAG][FAME]

− d[DAG]dt

= k2[DAG][MeOH] − k−2[MAG][FAME] − k1[TAG][MeOH] + k−1[DAG][FAME]

− d[MAG]dt

= k3[MAG][MeOH] − k−3[Gl][FAME] − k2[DAG][MeOH] + k−2[MAG][FAME]

[82]

Cottonseeda Ethanol 78.2 ± 2 Three consecutivereversible reactions

− d[TAG]dt

= k1[TAG][EtOH]3 − k−1[DAG][FAEE]

− d[DAG]dt

= k2[DAG][EtOH]3 − k−2[MAG][FAEE] − k1[TAG][EtOH]3 + k−1[DAG][FAEE]

− d[MAG]dt

= k3[MAG][EtOH]3 − k−3[Gl][FAEE] − k2[DAG][EtOH]3 + k−2[MAG][FAEE]

[167]

JatrophaMahua

Methanol 28 and 45 One step reaction(overall reaction)

− d[TAG]dt

= k[TAG]2[MeOH] [78]

Cottonseedb Methanol 60 One irreversiblereaction (overallreaction)

− d[TAG]dt

= k[TAG] [51]

Ethanol 80

Castor Ethanol 30–70 One irreversiblereaction (overallreaction)

− d[TAG]dt

= k[TAG] [86]

Jatropha–waste foodoil mixture

Methanol 50 Three consecutivereversible(methanolysis) + fivesaponificationreactions

d[TAG]dt

= [OH](−k1[TAG][MeOH] + k−1[DAG][FAME] − k5[TAG])

d[DAG]dt

= [OH](k1[TAG][MeOH] − k−1[DAG][FAME] − k2[DAG][MeOH] + k−2[MAG][FAME] + k5[TAG] − k6[DAG])

d[MAG]dt

= [OH](k2[DAG][MeOH] − k−2[MAG][FAME] − k3[MAG][MeOH] + k−3[Gl][FAME] + k6[DAG] − k7[MAG])

d[Gl]dt

= [OH](k3[MAG][MeOH] − k−3[Gl][FAME] + k7[MAG])

d[FAME]dt

= − d[MeOH]dt

= [OH](k1[TAG][MeOH] − k−1[DAG][FAME] + k2[DAG][MeOH] − k−2[MAG][FAME]

+k3[MAG][MeOH] − k−3[Gl][FAME] − k4[FAME])d[OH]

dt= − d[A]

dt= [OH](−k4[FAME] − k5[TAG] − k6[DAG] − k7[MAG] − k8[FFA])

d[W]dt

= − d[FFA]dt

= − k8[FFA][OH]

[76]

Pre-esterified jatrophaoil

Methanol 32–51 One irreversiblereaction (overallreaction)

− d[TAG]dt

= k[TAG]2 [136]

a In the presence of water.b Using both mechanical stirring and low frequency ultrasound.c [TAG], [DAG], [MAG], [FAME], [FAEE], [MeOH], [EtOH], [Gl], [FFA], [OH], [A], [W] – concentrations of TAG, DAG, MAG, FAME, FAEE, methanol, ethanol, glycerol, FFA, OH− , soap and water, respectively, t – time, k1, k2 and k3 –

forward alcoholysis reaction rate constant, k−1, k−2 and k−3 – reverse alcoholysis reaction rate constant, k – overall reaction rate constant, k4, k5, k6, k7, k8 – saponification reactions rate constants.

tainab

tewomoFmoaac

wcmtof

isybscs

wDiarwwoeketcVo

−

Ttoot7

oc

I.B. Bankovic-Ilic et al. / Renewable and Sus

he irreversible first-order kinetic model with respect to acylglyc-rols and its activation energy, in the temperature range 30–70 ◦C,as 70.6 kJ/mol [86]. Lu et al. [136] established a pseudo second-

rder kinetic model with respect to TAG for the base-catalyzedethanolysis of pre-esterified jatropha oil with activation energy

f 15.46 kJ/mol in the temperature range between 32 and 51 ◦C.or the base-catalyzed methanolysis of cottonseed oil, using bothechanical stirring and low frequency ultrasound, the first and sec-

nd order reactions with respect to TAG were reported, although better fit was obtained for the former [51]. Exceptionally, Jainnd Sharma [97] studied the kinetic of both steps in the acid-baseatalyzed transesterification of jatropha oil with respect to FAME:

d[FAME]dt

= k[FAME] (1)

here [FAME] is the FAME concentration, t is time and k is rateonstant. Based on the rate constant values for esterification andethanolysis reactions it was observed that the esterification reac-

ion was slower than the methanolysis. The activation energyf methanolysis reaction in the temperature range 35–60 ◦C wasound to be 87.81 kJ/mol [97].

The reaction mechanism of heterogeneously catalyzed reactionss more complex and includes various reactions that occur with theolid catalyst. The kinetics of heterogeneously catalyzed alcohol-sis has been rarely studied, and the various kinetic models haveeen reported. The kinetics of different vegetable oils methanoly-is over heteropolyacids supported on K-10 clay in the kineticallyontrolled region was described by three consecutive, reversible,econd-order reactions as follows [120]:

−d[TAG]dt

= w(k1[TAG][MeOH] − k−1[DAG][FAME])d[DAG]

dt= w(k1[TAG][MeOH] − k−1[DAG][FAME]

−k2[DAG][MeOH] + k−2[MAG][FAME])d[MAG]

dt= w(k2[DAG][MeOH] − k−2[MAG][FAME]

−k3[MAG][MeOH] + k−3[Gl][FAME])

(2)

here [MAG], [DAG] and [TAG] are the concentrations of MAG,AG and TAG, [MeOH] is the concentration of methanol, [FAME]

s the concentration of FAME, w is the catalyst loading, t is timend ki and k−i are rate constant of forward and reverse reactions,espectively (i = 1, 2 and 3). The experimental data were fitted wellith the kinetic model, but the kinetic parameters for all used oilsere not reported. The same model was applied for the jatropha

il methanolysis in the presence of KSF clay and Amberlyst 15. Thexperimental data were agreed well with the kinetic model, but noinetic parameters were published [110]. Vyas et al. [113] and Ziebat al. [123] established simpler kinetic models for jatropha and cas-or oil methanolysis over KNO3/Al2O3 and Zn5(OH)8(NO3)2·2H2Oatalyst, respectively. In order to fit the experimental data better,yas et al. [113] assumed that the overall reaction followed nthrder:

d[TAG]dt

= k[TAG]n (3)

he best fit was obtained for n ≈ 0.5. The activation energy for reac-ion carried out at 50–70 ◦C was estimated to be 112.8 kJ/mol. Theverall reaction rate of the castor oil methanolysis, due to the excessf methanol, followed the pseudo-first order kinetics with respecto TAG and the apparent activation energy was calculated to be

3.3 kJ/mol in the temperature range from 20 to 60 ◦C [123].

For the enzyme-catalyzed alcoholysis, only a limited numberf kinetic studies are found in the literature. Generally, the mostommonly used model to describe the kinetics of FAAE synthesis

le Energy Reviews 16 (2012) 3621– 3647 3641

using lipases as a catalyst is the Ping Pong Bi-Bi mechanism withcompetitive alcohol inhibition [168–171]:

vi = Vmax[TAG][A]Km,TAG[A](1 + [A]/Ki,A) + Km,A[TAG] + [TAG][A]

(4)

where are vi – initial rate, Vmax, Km,TAG, Km,A and Ki,A – kinetic con-stants, [TAG] and [A] – concentrations of TAG and acyl acceptor,respectively. Considering the interesterification of TAG and methylacetate for biodiesel production, Xu et al. [171] also developeda three-consecutive reversible-reaction kinetic model (shown byEq. (2)) and reported that the first step, i.e. reaction TAG to DAG,was the rate-limiting step for the overall reaction. Torres et al.[172,173] proposed the modified Michaelis–Menten mechanismthat incorporates or not the reaction reversibility and the enzymedeactivation, depending on the lipase source.

Up to date, the kinetics of enzyme-catalyzed non-edible oilsalcoholysis has hardly been studied. The kinetics of the castor oilethanolysis in supercritical carbon dioxide [162] and the cotton-seed oil methanolysis in the presence of ectoine [145] catalyzedby Novozym 435, was described by the Ping Pong Bi-Bi mecha-nism with competitive alcohol inhibition (Eq. (4)). By studying theeffect of temperature on the initial reaction rate, Royon et al. [142]calculated the lumped activation energy of 19 ± 2 kJ/mol that cor-responded to the forward first stepwise reaction of the cottonseedoil methanolysis.

Under supercritical conditions, the reaction mixture becomes asingle homogeneous phase without mass transfer limitations thataccelerates the reaction rate. The reaction rate of supercritical alco-holysis depends on the composition of the vegetable oil and theused alcohol. Oils containing TAG of saturated fatty acids reactfaster than oils with TAG of unsaturated acids, and the reactionbecomes slower with an increasing number of double bonds in theunsaturated acids [161,162]. The alcoholysis reaction rate of cas-tor oil containing a high level of ricinoleic acid, a hydroxy fattyacid, cannot be compared with those of other vegetable oils [162].Generally, the kinetic analyses of the supercritical alcoholysis up todate assumed one-step reaction and the pseudo first-order reactionkinetics with respect to TAG [156,157,159–162]:

−d[TAG]dt

= k[TAG] (5)

The activation energies for jatropha and pongamia oil methanolysisand ethanolysis at temperatures range 200–400 ◦C were calculatedto be between 9.1 and 11.4 kJ/mol [161], which were lower thanthose determined for the castor oil methanolysis and ethanolysis(35 and 55 kJ/mol, respectively) in almost the same temperatureinterval [162]. There was no obvious discontinuity near the criti-cal temperature. Tang et al. [156] established a pseudo first-orderreaction kinetics with respect to all glycerides species (TAG, DAG,MAG) and FFA for the jatropha oil methanolysis catalyzed by micro-NaOH with activation energy of 84.1 kJ/mol in the temperaturerange 250–275 ◦C. The jatropha oil transesterification via super-critical methyl acetate in the temperature range 300–345 ◦C wasdescribed by the pseudo first-order reaction kinetics with respectto TAG [157]. The higher value of calculated activation energy forthis process (364 kJ/mol) indicates that methyl acetate is less reac-tive in comparison with methanol or ethanol. Chen et al. [159,160]studied the kinetic of FAME synthesis using subcritical hydrolysisand subsequent supercritical methanolysis of jatropha oil extractedfrom seeds and kernels in the temperature range from 250–290 ◦C.The both steps follow the pseudo first-order reaction kinetics withrespect to TAG. For jatropha seed and kernel oil, the calculated acti-

vation energies for hydrolysis were 50.2 and 68.5 kJ/mol, and formethanolysis 23.9 and 45.2 kJ/mol, respectively. The higher valueof activation energy for the hydrolysis indicated that this reactionwas the rate determining step for biodiesel production, but the

3 tainab

dt

4t

isuncc

4

wpctFItbrtvo

fa[anidwiclTbuapce3epni[sf

eptifltwtTp

642 I.B. Bankovic-Ilic et al. / Renewable and Sus

ifference between the activation energies of methanolysis reac-ion of seed and kernel oil was not explained by the authors.

.3. The possibilities for improvement of non-edible oilsransesterification

The improvement of the alcoholysis reaction includes all activ-ties that increase the ester yield, shorten the reaction time and/orimplify the process. Several different novel methods have beensed in the last decade to improve the biodiesel production fromon-edible oils such as the application of in situ transesterifi-ation processes, ultrasonic irradiation, microwave heating ando-solvent.

.3.1. In situ processesUsing conventional methods, a non-edible oil is directly treated

ith an alcohol in the presence of a catalyst. Therefore, the totalrocess involves various oil treatments such as its extraction, purifi-ation, esterification and/or transesterification. The novel in situechnology for producing biodiesel from non-edible oils with highFA content involves only one step to finish all necessary stages.ncluding both extraction and esterification/transesterification,his technology is known as reactive extraction. The alcohol acts asoth an extracting solvent and an esterification/transesterificationeagent. Once the oil is extracted out from seeds, it is convertedo esters. In this way, it is possible to reduce processing time, sol-ent amount and total cost. However, it requires a higher amountf alcohol than traditional processes.

Several researchers have studied in situ biodiesel productionrom non-edible oil bearing materials in the absence of a cat-lyst [174] and the presence of an acid [101,164,175], a base51] or an enzyme [152,176] catalyst. A review of in situ cat-lytic and non-catalytic biodiesel production processes oil fromon-edible seeds is given in Table 11. The use of base catalysts

s more favorable than acid ones. Georgogianni et al. [51] pro-uced esters by base-catalyzed in situ process from cottonseedsith the oil containing a low FFA amount employing both mechan-

cal agitation and ultrasonication. In the presence of FFA in a highontent, the use of a base catalyst leads to the well-known prob-ems of soap formation and difficulty in product separation [164].he use of an acid catalyst in an in situ one-step process in theiodiesel production from jatropha seeds increases the ester yieldp to 99.8% [164,175], compared to the conventional two-stepcid/base [96], two-step enzymatic or in situ enzymatic [152,176]rocess. In the last case, the yield improvement by the lipase-atalyzed in situ reactive extraction with respect to the two-stepxtraction/lipase-catalyzed transesterification was 5.3% [176] or1.7% [152]. Although the required reaction time for the optimumster yield is in the range from 10 to 24 h, which is longer com-ared to conventional transesterification processes, it should beoted that in reactive extraction, both extraction and transester-

fication occur simultaneously. Unlike reports for jatropha seeds152,164,175,176], the conversion in the case of other vegetablepecies such as castor was lower for the reactive extraction thanor conventional methods [101].

The highest FAME yield was achieved from jatropha seeds bymploying non-catalytic supercritical reactive extraction at high-ressure conditions [174]. The 103.5% of FAME yield exceeds theheoretical one because of the excess of extracted oil. The co-solvents effective at the beginning of the process and the supercriticaluid extraction dominates at higher temperatures, which bringshe total oil extraction efficiency to 100%. The effect of co-solvent

as more evident in the in situ catalytic process than in the conven-

ional one in the cases of jatropha [175] and castor red [101] seeds.he co-solvent can facilitate the diffusion of the alcohol through thearticles of the seeds. However, this phenomenon was not observed

le Energy Reviews 16 (2012) 3621– 3647

with nordestina variety castor seeds, where the conversion withethanol is higher than that with the solvent mixture [101].

The use of ultrasonic energy improves in situ transesterifica-tion process in base-catalyzed biodiesel production from cotonseed[51]. The increase in esters yield was obtained in a remarkably shorttime (20 min and 40 min in reactions with methanol and ethanol,respectively), and the yields of the isolated esters were higherfor the ultrasonicated reaction system than for the mechanicallystirred system [51].

The efficiency of oil extraction from seeds of different sizeincreases gradually with the progress of the reactive extraction pro-cess [175]. The size of seeds is a significant factor affecting the esteryield. Higher oil extraction efficiencies were achieved from smallerseeds, due to a larger contact surface area and a less hindrance fromthe outer shell covering the oil seeds [174,175]. The oil extractionfrom the seeds is favor until the efficiency reaches more than 90%.For larger seeds, there is a stage where mass transfer limitationexists. Also, they settle-down easer and reduce the methanol:oilcontact surface area for the transesterification process.

4.3.2. The use of ultrasonic irradiationThe use of ultrasonic irradiation is a new, more efficient mixing

method in biodiesel production, compared to conventional meth-ods of agitation. The ultrasonic energy improves the mass transferbetween the immiscible reactants, increases the chemical reactionrate and ester yield and decreases the reaction time and energy con-sumption. Deng et al. [77] concluded that two-step process coupledwith ultrasonic irradiation is an efficient method for biodiesel pro-duction from crude jatropha oil with high FFA value. They showedthat the high FAME yield (96.4%) can be achieved in only 30 min(Table 5). Ultrasound-assisted base catalyzed ethanolysis of cotton-seed oil under optimal reaction conditions gives 90% ester yield in60 min, while in the case of mechanical agitation only 86% esteryield is obtained for the same time [51] (Table 4). This can beexplained assuming that soap formation is reduced in the presenceof ultrasound action [51]. Ultrasonic irradiation is also effective inincreasing the ester yield in the in situ biodiesel production fromcottonseeds [51] (Table 11) and race bran [177].

4.3.3. The use of microwave heatingMicrowave irradiation is an efficient heating method used to

reduce the reaction time and to improve the ester yield in thetransesterification reaction of non-edible oils. Its advantages aresimplicity, low cost and energy economy [50,99,127]. Heat is gen-erated during this process by molecular friction, because polaralcohol molecules align with the continuously changing magneticfield generated by microwaves. The increase of the reaction rate isattributed to an enhanced temperature at the catalytic surface [50].Because of the effective heat transfer, processes with microwaveirradiation can be completed in a much shorter time than thosewith conventional heating [50,99]. Microwave irradiation seems tohave a strong effect on the reaction rate rather than on the esteryield [99]. Comparing the base catalyzed methanolysis of cotton-seed oil under microwave and conventional heating at 60 ◦C, Azcanand Danisman [50] observed approximately the same ester yieldsin both processes in 7 min and 30 min, respectively. The reactionrate of microwave-assisted reaction of karanja oil with methanolincreases drastically, which reduces the reaction time to 2–3 min,compared to 1–2 h in the case of conventional heating [178].Identical FAME yields are achieved by homogeneously and het-erogeneously acid-catalyzed transesterification under microwaveradiation [126]. There are contradictory conclusions on the reac-

tivity of methanol and ethanol under microwave irradiation. WhileYuan et al. [126] observed that the methanol system achieved theequilibrium in 1 h and was twice as fast as the ethanol system.Perin et al. [125] achieved the same biodiesel yield in a shorter

I.B. Bankovic-Ilic

et al.

/ R

enewable

and Sustainable

Energy R

eviews

16 (2012) 3621– 36473643

Table 11A review of the in situ transesterification of different non-edible feedstocks.

Feedstock (oil) Type, volume of reactor, cm3/Type ofagitator, agitation intensity, rpm

Acyl acceptor Alcohol to seedratio, mL/g

Catalyst/loading,wt.% to the seed

Temperature, ◦C Optimal reaction conditions Reference

Reactionconditions

Oil extractionefficiency/yield(conversion),%/Time, h

Jatropha Flask, 250/Magnetic, – Methanol 7.5 H2SO4/1 60 91.2/99.8a/24 [175]Flask, 500/Magnetic, – Methanol 5–20 H2SO4/5–30 30–60 60 ◦C, 10.5 mL/g,

21% H2SO4

98.1/10 [164]

Batch reactor, 450/Mechanical, 400 Methanol 10 – 200–300 300 ◦C 103.5b/24 [174]Screw-caped glass vials/–, 180 Methyl-acetate 5–20 Novozym 435/30 50 7.5 mL/g 86.1/24 [176]

Ethyl-acetate 87.2/24Screw-caped glass vials/–, 180 Dimethyl

carbonate5–20 Novozym 435/10 50 10 mL/g 95.9/24 [152]

Diethyl carbonate94.5/24

Castor Flask, 100/Mechanical, 300 I step: Ethanol 40:1c H2SO4/1 60 – [101]II step: Ethanol 20:1c KOH/1 60 (51.2)/1 and

(65.6)d/1Cottonseed Batch reactor, 500/Mechanical, 600

and ultrasoundMethanol 7:1 NaOH/2 60 97/2, 97/0.67e [51]

Ethanol 7:1 NaOH/2 80 78/6, 98/0.67e

Rice bran Methanol 10:1 H2SO4/27.6 60 82.79f/6 (mediumFFA content of47.87%)79.08/6 (low FFAcontent of 13.27%)

[177]

a n-hexane as co-solvent (10% vol. of methanol).b supercritical conditions: 240 MPa; n-hexane as co-solvent (2.5 mL/g of seed).c alcohol:oil molar ratio.d n-hexane as co-solvent (20%, v/v to oil).e ultrasonic power: 200 W; frequency: 24 kHz.f ultrasonic power: 500 W; frequency: 35 kHz.

3 tainab

tivaatl

4

cmostspedccogdFiiaobartmmvpc

5

sonifhBgtb(ictceetawge

644 I.B. Bankovic-Ilic et al. / Renewable and Sus

ime with ethanol than with methanol. A continuous microwaverradiation reactor was developed by Liao and Chung [98] to con-ert jatropha oil to alkyl ester. Since the microwave power was sett 80 W and the energy consumption in the 10 min process wasbout 48 kJ, it was concluded that the ratio of energy consumptiono the biodiesel generation would be decreased significantly in thearge-scale microwave system.

.3.4. The effect of co-solventThe use of a co-solvent in the reaction mixture improves mis-

ibility of vegetable oil and alcohol differing in polarities [179],inimizes the induction period [180] and enables the system to

vercome the initial mass transfer resistance. The added solventhould be inert and cannot either react with reactants or deactivatehe catalyst. By adding a co-solvent to the reaction mixture, theolubility of alcohol in the oil phase increases, an oil-rich single-hase is generated and the reaction rate and the ester yield arenhanced. Tetrahydrofuran (THF), methyl tert-butyl ether (MTBE),imethyl ether, diethyl ether and n-hexane are the most often usedo-solvents in the biodiesel production. The THF and n-hexane arehosen as co-solvents because of their boiling points close to thatf methanol and their easy recovery at the reaction end. THF is aood co-solvent because of its miscibility with methanol and waterue to hydrogen bonding. In the presence of THF as co-solvent, theAME yield in the base catalyzed transesterification of karanja oil isncreased up to 95%, compared that of 92% achieved in its absencen the same time [81]. An almost complete conversion (99%) waschieved with THF in the case of mahua oil [78]. However, the usef THF in the case of jatropha oil transesterification appears not toe necessary, especially because of the cost involved in the sep-ration of THF from the reaction mixture. Due to a high reactionate of jatropha oil, the rate-enhancing effect of THF is less thanhat in the case of mahua oil [78]. The use of co-solvent is recom-

ended for the reactions performed at lower temperatures whenass transfer effects dominate [78]. When ethanol was used as sol-

ent and n-hexane as co-solvent in the two-step transesterificationrocess of castor oil, it was not observed a significant effect on theonversion and the reaction rate [101].

. Conclusions

Because of environmental pollution by fuel combustion emis-ion, an increase of the world energy demand and competitionf edible oil sources for human use and biofuel production, theon-edible oils have become the leading raw materials for obtain-

ng biodiesel. Of several possible methods for biodiesel productionrom non-edible oils, their transesterification reaction with an alco-ol in the presence of a catalyst is the most suitable method.oth one- and two-step processes of homogeneously and hetero-eneously catalyzed alcoholysis of non-edible oils are used today inhe biodiesel synthesis. In the case of one-step processes, the choiceetween base and acid catalysts mainly depends on the FFA contentor acid value) in the non-edible oil. Base catalysts are preferablen the case of non-edible oils with a lower FFA content, while acidatalysts are suitable to the oils having a high amount of FFAs dueo the ability of acid catalyst to simultaneously catalyze esterifi-ation and alcoholysis reaction. Two-step (acid/base) processes,mployed when non-edible oils have a high FFA content, are anffective method to obtain a high biodiesel yield within a short reac-ion time. Sometimes, the modified catalysts with dual basic and

cidic sites or the mixture of acid and base catalysts are employedith the same goal. The use of solid catalysts instead of homo-

eneous ones is more desired for many reasons such as reducednvironmental pollution and higher qualities of biodiesel and

le Energy Reviews 16 (2012) 3621– 3647

glycerol. However, a completely heterogeneous two-step processhas not been developed yet.

The use of immobilized lipases and supercritical conditions fortransesterification of non-edible oils are currently among the mostpromising methods for biodiesel production. Lipases are especiallysuitable because they can catalyze simultaneously both TAG alco-holysis and FFA esterification. The carrying out in non-aqueousenvironments, easy recovery of glycerol and the use of feedstockwith high FFA content are also the advantages of applying enzymesfor biodiesel synthesis from non-edible oils over the other meth-ods. However, the relatively high price of lipases at present is theirmain drawback for their use on the commercial scale. Compared toconventional methods, the transesterification reaction rate undersupercritical conditions is higher, the reaction time is shorter, sep-aration and purification of biodiesel are simply, there is no soapformation and no waste generation from catalyst separation andabsence of catalyst makes the glycerol recovery easier.

The use of statistical methods, providing the interactionsbetween variables optimizing the process conditions and allowingbetter knowledge of the process, contributes to the development ofeconomically efficient processes of biodiesel production based onnon-edible plant oils. Mathematical models describing the reactionrate are necessary for understanding the fundamental transester-ification kinetics for the reactor design and process scale-up. Notrequiring complex computation, simple kinetic models are favor-able when fitted well the experimental data. Several different novelmethods have been used in the last decade to improve the biodieselproduction from non-edible oils. These methods such as the appli-cation of in situ transesterification processes, ultrasonic irradiation,microwave heating and co-solvent reduce processing time andenergy consumption and enhance the ester yield.

Acknowledgement

This work has been funded by the Ministry of Education andScience of the Republic of Serbia (Project III 45001).

References

[1] Balat M, Balat H. Progress in biodiesel processing. Appl Energy2010;87:1815–35.

[2] Moser BR. Biodiesel production, properties, and feedstocks. In Vitro Cell DevPl 2009;45:229–66.

[3] Zhang Y, Dube MA, McLean DD, Kates M. Biodiesel production from wastecooking oil: 1. Process design and technological assessment. Bioresour Tech-nol 2003;89:1–16.

[4] Balat M. Potential alternatives to edible oils for biodiesel production – areview. Energy Convers Manage 2011;52:1479–92.

[5] Fan X, Burton R. Recent development of biodiesel feedstocks and the applica-tions of glycerol: a review. Open Fuels Energy Sci J 2009;2:100–9.

[6] Gui MM, Lee KT, Bhatia S. Feasibility of edible oil vs. non-edible oil vs. wasteedible oil as biodiesel feedstock. Energy 2008;33:1646–53.

[7] Leung DYC, Wu X, Leung MKH. A review on biodiesel production using cat-alyzed transesterification. Appl Energy 2010;87:1083–95.

[8] Karmakar A, Karmakar S, Mukherjee S. Properties of various plantsand animals feedstocks for biodiesel production. Bioresour Technol2010;101:7201–10.

[9] Demirbas A. Progress and recent trends in biodiesel fuels. Energy ConversManage 2009;50:14–34.

[10] Kralova I, Sjöblom J. Biofuels – renewable energy sources: a review. J DispersSci Technol 2010;31:409–25.

[11] Yusuf NNAN, Kamarudin SK, Yaakub Z. Overview on the current trends inbiodiesel production. Energy Convers Manage 2011;52:2741–51.

[12] Kumar A, Sharma S. Potential non-edible oil resources as biodiesel feedstock:an Indian perspective. Renew Sustain Energy Rev 2011;15:1791–800.

[13] Atadashi IM, Aroua MK, Aziz AA. High quality biodiesel and its diesel engineapplication: a review. Renew Sustain Energy Rev 2010;14:1999–2008.

[14] Basha SA, Gopal KR, Jebaraj S. A review on biodiesel production, combustion,emissions and performance. Renew Sustain Energy Rev 2009;13:1628–34.

[15] Hasheminejad M, Tabatabaei M, Mansourpanah Y, Khatami M, Javani A.Upstream and downstream strategies to economize biodiesel production.Bioresour Technol 2011;102:461–8.

[16] No SY. Inedible vegetable oils and their derivatives for alternative diesel fuelsin CI engines: a review. Renew Sustain Energy Rev 2011;15:131–49.

tainab

I.B. Bankovic-Ilic et al. / Renewable and Sus

[17] Helwani Z, Othman MR, Aziz N, Fernando WJN, Kim J. Technologies for pro-duction of biodiesel focusing on green catalytic techniques: a review. FuelProcess Technol 2009;90:1502–14.

[18] Juan JC, Kartika DA, Wub TY, Hin TYY. Biodiesel production from jatropha oilby catalytic and non-catalytic approaches: an overview. Bioresour Technol2011;102:452–60.

[19] Koh MY, Ghazi TIM. A review of biodiesel production from Jatropha curcas L.oil. Renew Sustain Energy Rev 2011;15:2240–51.

[20] Shahid EM, Jamal Y. Production of biodiesel: a technical review. Renew Sus-tain Energy Rev 2011;15:4732–45.

[21] Vyas AP, Verma JL, Subrahmanyam N. A review on FAME production pro-cesses. Fuel 2010;89:1–9.

[22] Janaun J, Ellis N. Perspectives on biodiesel as a sustainable fuel. Renew SustainEnergy Rev 2010;14:1312–20.

[23] Jain S, Sharma MP. Prospect of biodiesel from Jatropha in India: a review.Renew Sustain Energy Rev 2010;14:763–71.

[24] Kumar S, Chaubea A, Jain SK. Sustainability issues for promotion of Jat-ropha biodiesel in Indian scenario: a review. Renew Sustain Energy Rev2012;16:1089–98.

[25] Silitonga AS, Atabania AE, Mahliaa TMI, Masjukia HH, Badruddina IA,Mekhilefe S. A review on prospect of Jatropha curcas for biodiesel in Indonesia.Renew Sustain Energy Rev 2011;15:3733–56.

[26] Murugesan A, Umarani C, Chinnusamy TR, Krishnan M. Production and anal-ysis of biodiesel from non-edible oils – a review. Renew Sustain Energy Rev2009;13:825–34.

[27] Chisti Y. Biodiesel from microalgae. Biotechnol Adv 2007;25:294–306.[28] Searchinger T, Heimlich R, Houghton RA, Dong F, Elobeid A, Fabiosa J, et al. Use

of U.S. croplands for biofuels increases greenhouse gases through emissionsfrom land-use change. Science 2008;319:1238–40.

[29] Azam MM, Waris A, Nahar NM. Prospects and potential of fatty acid methylesters of some non-traditional seed oils for use as biodiesel in India. BiomassBioenergy 2005;29:293–302.

[30] Wen Z, Johnson MB. Microalgae as a feedstock for biodiesel production. Com-munications and marketing. Blacksburg, VA: College of Agriculture and LifeSciences, Virginia Polytechnic Institute and State University, Publication 442-886; 2009.

[31] González AR, Gallego EG. Operation variables in the animal-fat transesterifi-cation process: a review. Ing Univ 2011;15:197–218.

[32] Phan AN, Phan TM. Biodiesel production from waste cooking oils. Fuel2008;87:3490–6.

[33] Enweremadu CC, Mbarawa MM. Technical aspects of production and analysisof biodiesel from used cooking oil – a review. Renew Sustain Energy Rev2009;13:2205–24.

[34] Kulkarni MG, Dalai AK. Waste cooking oil – an economical source for biodiesel:a review. Ind Eng Chem Res 2006;45:2901–13.

[35] Meher LC, Sagar DV, Naik SN. Technical aspects of biodiesel production bytransesterification – a review. Renew Sustain Energy Rev 2006;10:248–68.

[36] Refaat AA. Different techniques for the production of biodiesel from wastevegetable oil. Int J Environ Sci Technol 2010;7:183–213.

[37] Amaro HM, Guedes AC, Malcata FX. Advances and perspectives in usingmicroalgae to produce biodiesel. Appl Energy 2011;88:3402–10.

[38] Patil P, Gude VG, Deng S. Biodiesel production from Jatropha curcas, wastecooking and Camelina sativa oils. Ind Eng Chem Res 2009;48:10850–6.

[39] Tiwari AK, Kumar A, Raheman H. Biodiesel production from jatropha oil(Jatropha curcas) with high free fatty acids: an optimized process. BiomassBioenergy 2007;31:569–75.

[40] Naik M, Meher LC, Naik SN, Das LM. Production of biodiesel from high freefatty acid karanja (Pongamia pinnata) oil. Biomass Bioenergy 2008;32:354–7.

[41] Giannelos PN, Zannikos F, Stournas S, Lois E, Anastopoulos G. Tobacco seed oilas an alternative diesel fuel: physical and chemical properties. Ind Crop Prod2002;16:1–9.

[42] Usta N. Use of tobacco seed oil methyl ester in a turbocharged indirect injec-tion diesel engine. Biomass Bioenergy 2005;28:77–86.

[43] Veljkovic VB, Lakicevic SH, Stamenkovic OS, Todorovic ZB, Lazic ML. Biodieselproduction from tobacco (Nicotiana tabacum L.) seed oil with a high contentof free fatty acids. Fuel 2006;85:2671–5.

[44] Ghadge SV, Raheman H. Biodiesel production from mahua (Madhuca indica)oil having high free fatty acids. Biomass Bioenergy 2005;28:601–5.

[45] Jena PC, Raheman H, Kumar GVP, Machavaram R. Biodiesel production frommixture of mahua and simarouba oils with high free fatty acids. BiomassBioenergy 2010;34:1108–16.

[46] Nabi MdN, Hustad JE, Kannan D. First generation biodiesel production fromnon-edible vegetable oil and its effect on diesel emissions. In: Proceedings ofthe 4th BSME–ASME international conference on thermal engineering. 2008.

[47] Ramadhas AS, Jayaraj S, Muraleedharan C. Biodiesel production from high FFArubber seed oil. Fuel 2005;84:335–40.

[48] Kansedo J, Lee KT, Bhatia S. Cerbera odollam (sea mango) oil as a promisingnon-edible feedstock for biodiesel production. Fuel 2009;88:1148–50.

[49] de Oliveira D, Di Luccio M, Faccio C, Rosa CD, Bender JP, Lipke N, et al. Optimiza-tion of enzymatic production of biodiesel from castor oil in organic solventmedium. Appl Biochem Biotechnol 2004;113-116:771–80.

[50] Azcan N, Danisman A. Alkali catalyzed transesterification of cottonseed oil bymicrowave irradiation. Fuel 2007;86:2639–44.

[51] Georgogianni KG, Kontominas MG, Pomonis PJ, Avlonitis D, Gergis V. Alkalineconventional and in situ transesterification of cottonseed oil for the produc-tion of biodiesel. Energy Fuels 2008;22:2110–5.

le Energy Reviews 16 (2012) 3621– 3647 3645

[52] Sharma YC, Singh B. An ideal feedstock, kusum (Schleichera triguga) for prepa-ration of biodiesel: optimization of parameters. Fuel 2010;89:1470–4.

[53] Karaosmanoglu F, Tüter M, Göllü E, Yanmaz S, Altintig E. Fuel properties ofcottonseed oil. Energy Sources 1999;21:821–8.

[54] Usta N, Aydogan B, C on AH, Uguzdogan E, Özkal SG. Properties and quality ver-ification of biodiesel produced from tobacco seed oil. Energy Convers Manage2011;52:2031–9.

[55] Singh D, Singh SP. Low cost production of ester from non edible oil of Argemonemexicana. Biomass Bioenergy 2010;34:545–9.

[56] Wang R, Hanna MA, Zhou WW, Bhadury PS, Chen Q, Song BA, et al. Pro-duction and selected fuel properties of biodiesel from promising non-edibleoils: Euphorbia lathyris L., Sapium sebiferum L. and Jatropha curcas L. BioresourTechnol 2011;102:1194–9.

[57] Li X, He XY, Li ZL, Wang YD, Wang CY, Shi H, et al. Enzymatic production ofbiodiesel from Pistacia chinensis bge seed oil using immobilized lipase. Fuel2012;92:89–93.

[58] Wang R, Zhou W-W, Hanna MA, Zhang Y-P, Bhadury PS, Wang Y, et al.Biodiesel preparation, optimization, and fuel properties from non-ediblefeedstock, Datura stramonium L. Fuel 2012;91:182–6.

[59] Sirisomboon P, Kitchaiya P, Pholpho T, Mahuttanyavanitch W. Physical andmechanical properties of Jatropha curcas L. fruits, nuts and kernels. BiosystEng 2007;97:201–7.

[60] Meneghetti SMP, Meneghetti MR, Wolf CR, Silva EC, Lima GES, Silva LL, et al.Biodiesel from castor oil: a comparison of ethanolysis versus methanolysis.Energy Fuels 2006;20:2262–5.

[61] Demirbas A. Studies on cottonseed oil biodiesel prepared in non-catalytic SCFconditions. Bioresour Technol 2008;99:1125–30.

[62] Kafuku G, Mbarawa M. Alkaline catalyzed biodiesel production fromMoringa oleifera oil with optimized production parameters. Appl Energy2010;87:2561–5.

[63] Kafuku C, Lam MK, Kansedo J, Lee KT, Mbarawa M. Heterogeneous cat-alyzed biodiesel production from Moringa oleifera oil. Fuel Process Technol2010;91:1525–9.

[64] Rashid U, Anwar F, Moser BR, Knothe G. Moringa oleifera oil: a possible sourceof biodiesel. Bioresour Technol 2008;99:8175–9.

[65] Deka DC, Basumatary S. High quality biodiesel from yellow oleander (Thevetiaperuviana) seed oil. Biomass Bioenergy 2011;35:1797–803.

[66] Moser BR, Knothe G, Vaughn SF, Isbell TA. Production and evaluationof biodiesel from field pennycress (Thlaspi arvense L.). Oil, Energy Fuel2009;23:4149–55.

[67] Ma F, Hanna MA. Biodiesel production: a review. Bioresour Technol1999;70:1–15.

[68] Srivastava A, Prasad R. Tryglicerides-based diesel fuels. Renew Sustain EnergyRev 2000;4:111–33.

[69] Ilham Z, Saka S. Two-step supercritical dimethyl carbonate methodfor biodiesel production from Jatropha curcas oil. Bioresour Technol2010;101:2735–40.

[70] Achten WMJ, Verchot L, Franken YJ, Mathijs E, Singh VP, Aerts R, et al. Jatrophabio-diesel production and use. Biomass Bioenergy 2008;32:1063–84.

[71] Chitra P, Venkatachalam P, Sampathrajan A. Optimisation of experimentalconditions for biodiesel production from alkali-catalysed transesterificationof Jatropha curcus oil. Energy Sustain Dev 2005;9:13–8.

[72] Berchmans HJ, Hirata S. Biodiesel production from crude Jatropha cur-cas L. seed oil with a high content of free fatty acids. Bioresour Technol2008;99:1716–21.

[73] Tapanes NCO, Aranda DAG, Carneiro JWDM, Antunes OAC. Transesterifica-tion of Jatropha curcas oil glycerides: theoretical and experimental studies ofbiodiesel reaction. Fuel 2008;87:2286–95.

[74] Zahira Y, Ong BH, Kumar SMN, Kamarudin SK. Microwave-assisted trans-esterification of jatropha and waste frying palm oil. Int J Sustain Energy2009;28:195–201.

[75] de Oliveira JS, Leite PM, de Souza LB, Mello VM, Silva EC, Rubim JC, et al.Characteristics and composition of Jatropha gossypiifolia and Jatropha cur-cas L. oils and application for biodiesel production. Biomass Bioenergy2009;33:449–53.

[76] Berchmans HJ, Morishita K, Takarada T. Kinetic study of hydroxide-catalyzedmethanolysis of Jatropha curcas–waste food oil mixture for biodiesel produc-tion. Fuel 2010, doi:10.1016/j.fuel.2010.01.017.

[77] Deng X, Fang Z, Liu Y. Ultrasonic transesterification of Jatropha curcas L.oil to biodiesel by a two-step process. Energy Convers Manage 2010;51:2802–7.

[78] Kumar GR, Ravi R, Chadha A. Kinetic studies of base-catalyzed transesterifica-tion reactions of non-edible oils to prepare biodiesel: the effect of co-solventand temperature. Energy Fuels 2011;25:2826–32.

[79] Kumar D, Kumar G, Johari R, Kumar P. Fast, easy ethanomethanolysis of Jat-ropha curcus oil for biodiesel production due to the better solubility of oil withethanol in reaction mixture assisted by ultrasonication. Ultrason Sonochem;in press.

[80] Sahu G, Das LM, Sharma BK, Naik SN. Pilot plant study on biodiesel productionfrom Karanja and Jatropha oils. Asia-Pac J Chem Eng 2011;6:38–43.

[81] Karmee SK, Chadha A. Preparation of biodiesel from crude oil of Pongamia

pinnata. Bioresour Technol 2005;96:1425–9.

[82] Karmee SK, Chandna D, Ravib R, Chadha A. Kinetics of base-catalyzedtransesterification of triglycerides from Pongamia oil. J Am Oil Chem Soc2006;83:873–7.

3 tainab

646 I.B. Bankovic-Ilic et al. / Renewable and Sus

[83] Ragit SS, Mohapatra SK, Kundu K, Prashant Gill. Optimization of neem methylester from transesterification process and fuel characterization as a dieselsubstitute. Biomass Bioenergy 2011;35:1138–44.

[84] Parlak A, Karabas H, Ayhan V, Yasar H, Soyhan HS, Ozsert I. Comparison ofthe variables affecting the yield of tobacco seed oil methyl ester for KOH andNaOH catalysts. Energy Fuels 2009;23:1818–24.

[85] da Silva NL, Maciel MRW, Batistella CB, Filho RM. Optimization of biodieselproduction from castor oil. Appl Biochem Biotechnol 2006;129:405–14.

[86] da Silva NL, Batistella CB, Filho RM, Maciel MRW. Biodiesel produc-tion from castor oil: optimization of alkaline ethanolysis. Energy Fuels2009;23:5636–42.

[87] Meneghetti SMP, Meneghetti MR, Serra TM, Barbosa DC, Wolf CR. Biodieselproduction from vegetable oil mixtures: cottonseed, soybean, and castor oils.Energy Fuels 2007;21:3746–7.

[88] Barbosa DC, Serra TM, Meneghetti SMP, Meneghetti MR. Biodiesel productionby ethanolysis of mixed castor and soybean oils. Fuel 2010;89:3791–4.

[89] Keera ST, El Sabagh SM, Taman AR. Transesterification of vegetable oil tobiodiesel fuel using alkaline catalyst. Fuel 2011;90:42–7.

[90] Saravanan N, Puhan S, Nagarajan G, Vedaraman N. An experimental compari-son of transesterification process with different alcohols using acid catalysts.Biomass Bioenergy 2010;34:999–1005.

[91] Valle PWPA, Rezende TF, Souza RA, Fortes ICP, Pasa VMD. Fractional facto-rial and Doehlert experimental designs in biodiesel production: ethanolysisof Raphanus sativus L. var. oleiferus stokes oil catalyzed by sodium ethoxide.Energy Fuels 2009;23:5219–27.

[92] Ma F, Clements LD, Hanna MA. Biodiesel fuel from animal fat. Ancillary studieson transesterification of beef talow. Ind Eng Chem Res 1998;37:3768–71.

[93] Freedman B, Pryde EH, Mounts TL. Variables affecting the yields of fatty estersfrom transesterified vegetable oils. J Am Oil Chem Soc 1984;61:1638–43.

[94] Van Gerpen J, Shanks B, Pruszko R, Clements D, Knothe G. Biodiesel productiontechnology. Subcontractor report. National Renewable Energy Laboratory,NREL/SR-510-36244; 2004.

[95] Cavalcante KSB, Penha MNC, Mendonc a KKM, Louzeiro HC, Vasconcelos ACS,Maciel AP, et al. Optimization of transesterification of castor oil with ethanolusing a central composite rotatable design (CCRD). Fuel 2010;89:1172–6.

[96] Patil P, Deng S. Optimization of biodiesel production from edible and non-edible vegetable oils. Fuel 2009;88:1302–6.

[97] Jain S, Sharma MP. Biodiesel production from Jatropha curcas oil. Renew Sus-tain Energy Rev 2010;14:3140–7.

[98] Liao CC, Chung TW. Analysis of parameters and interaction between param-eters of the microwave-assisted continuous transesterification process ofJatropha oil using response surface methodology. Chem Eng Res Des2011;89:2575–81.

[99] Kamath HV, Regupathi I, Saidutta MB. Optimization of two step karanjabiodiesel synthesis under microwave irradiation. Fuel Process Technol2011;92:100–5.

[100] Lakshmi CV, Viswanath K, Venkateshwar S, Satyavathi B. Mixing character-istics of the oil–methanol system in the production of biodiesel using edibleand non-edible oils. Fuel Process Technol 2011;92:1411–7.

[101] Hincapié G, Mondragón F, López D. Conventional and in situ transesterifica-tion of castor seed oil for biodiesel production. Fuel 2011;90:1618–23.

[102] Yusup S, Khan MA. Base catalyzed transesterification of acid treated veg-etable oil blend for biodiesel production. Biomass Bioenergy 2010;34:1500–4.

[103] D’Oca MGM, Haertel PL, de Moraes DC, Callegaro FJP, Kurz MHS, Primel EG,et al. Base/acid-catalyzed FAEE production from hydroxylated vegetable oils.Fuel 2011;90:912–6.

[104] Dossin TF, Reyniers M-F, Berger RJ, Marin GB. Simulation of heterogeneouslyMgO-catalyzed transesterification for fine-chemical and biodiesel industrialproduction. Appl Catal B: Environ 2006;67:136–48.

[105] Mbaraka IK, Shanks BH. Conversion of oils and fats using advanced meso-porous heterogeneous catalysts. J Am Oil Chem Soc 2006;83:79–91.

[106] Boey PL, Maniam GP, Hamid SA, Ali DMH. Utilization of waste cockle shell(Anadara granosa) in biodiesel production from palm olein: optimizationusing response surface methodology. Fuel 2011;90:2353–8.

[107] Boey PL, Maniam GP, Hamid SA. Biodiesel production via transesterificationof palm olein using waste mud crab (Scylla serrata) shell as a heterogeneouscatalyst. Bioresour Technol 2009;100:6362–8.

[108] Wei Z, Xu C, Li B. Application of waste eggshell as low-cost solid catalyst forbiodiesel production. Bioresour Technol 2009;100:2883–5.

[109] Viriya-empikul N, Krasae P, Puttasawat B, Yoosuk B, Chollacoop N, Faung-nawakij K. Waste shells of mollusk and egg as biodiesel production catalysts.Bioresour Technol 2010;101:3765–7.

[110] Zanette AF, Barella RA, Pergher SBC, Treichel H, Oliveira D, Mazutti MA, et al.Screening, optimization and kinetics of Jatropha curcas oil transesterificationwith heterogeneous catalysts. Renew Energy 2011;36:726–31.

[111] Supamathanon N, Wittayakun J, Prayoonpokarach S. Properties of Jat-ropha seed oil from Northeastern Thailand and its transesterificationcatalyzed by potassium supported on NaY zeolite. J Ind Eng Chem 2011;17:182–5.

[112] Taufiq-Yap YH, Lee HV, Hussein MZ, Yunus R. Calcium-based mixed oxide cat-

alysts for methanolysis of Jatropha curcas oil to biodiesel. Biomass Bioenergy2011;35:827–34.

[113] Vyas AP, Subrahmanyam N, Patel PA. Production of biodiesel throughtransesterification of Jatropha oil using KNO3/Al2O3 solid catalyst. Fuel2009;88:625–8.

le Energy Reviews 16 (2012) 3621– 3647

[114] Sree R, Babu NS, Prasad PSS, Lingaiah N. Transesterification of edibleand non-edible oils over basic solid Mg/Zr catalysts. Fuel Process Technol2009;90:152–7.

[115] Olutoye MA, Hameed BH. Synthesis of fatty acid methyl ester from crude jat-ropha (Jatropha curcas Linnaeus) oil using aluminium oxide modified Mg–Znheterogeneous catalyst. Bioresour Technol 2011;102:6392–8.

[116] Zhu H, Wu Z, Chen Y, Zhang P, Duan S, Liu X, et al. Preparation of biodieselcatalyzed by solid super base of calcium oxide and its refining process. ChinJ Catal 2006;27:391–6.

[117] Endalew AK, Kiros Y, Zanzi R. Heterogeneous catalysis for biodiesel produc-tion from Jatropha curcas oil (JCO). Energy 2011;36:2693–700.

[118] Babu NS, Sree R, Prasad PSS, Lingaiah N. Room-temperature transesterifica-tion of edible and nonedible oils using a heterogeneous strong basic Mg/Lacatalyst. Energy Fuels 2008;22:1965–71.

[119] Kumar D, Kumar G, Singh PCP. Ultrasonic-assisted transesterificationof Jatropha curcas oil using solid catalyst Na/SiO2. Ultrason Sonochem2010;17:839–44.

[120] Bokade VV, Yadav GD. Transesterification of edible and nonedible vegetableoils with alcohols over heteropolyacids supported on acid-treated clay. IndEng Chem Res 2009;48:9408–15.

[121] Kaur M, Ali A. Lithium ion impregnated calcium oxide as nano catalystfor the biodiesel production from karanja and jatropha oils. Renew Energy2011;36:2866–71.

[122] Meher LC, Kulkarni MG, Dalai AK, Naik SN. Transesterification of karanja(Pongamia pinnata) oil by solid basic catalysts. Eur J Lipid Sci Technol2006;108:389–97.

[123] Zieba A, Pacuza A, Drelinkiewicz A. Transesterification of triglycerides withmethanol catalyzed by heterogeneous zinc hydroxy nitrate catalyst. Evalua-tion of variables affecting the activity and stability of catalyst. Energy Fuels2010;24:634–45.

[124] Almeida RMD, Noda LK, Gonc alves NS, Meneghetti SMP, Meneghetti MR.Transesterification reaction of vegetable oils, using superacid sulfated TiO2-base catalysts. Appl Catal A: Gen 2008;347:100–5.

[125] Perin G, Álvaro G, Westphal E, Viana LH, Jacob RG, Lenardão EJ, et al.Transesterification of castor oil assisted by microwave irradiation. Fuel2008;87:2838–41.

[126] Yuan H, Yang BL, Zhu GL. Synthesis of biodiesel using microwave absorptioncatalysts. Energy Fuels 2009;23:548–52.

[127] Yuan H, Yang B, Zhang H, Zhou X. Synthesis of biodiesel using castor oil undermicrowave radiation. Int J Chem React Eng 2011;9. Article A71.

[128] Chen H, Peng B, Wang D, Wang J. Biodiesel production by the transesterifica-tion of cottonseed oil by solid acid catalysts. Front Chem Sci Eng 2007;1:11–5.

[129] Lingfeng C, Guomin X, Bo X, Guangyuan T. Transesterification of cottonseedoil to biodiesel by using heterogeneous solid basic catalysts. Energy Fuels2007;21:3740–3.

[130] Kumar D, Ali A. Nanocrystalline lithium ion impregnated calcium oxide as het-erogeneous catalyst for transesterification of high moisture containing cottonseed oil. Energy Fuels 2010;24:2091–7.

[131] Peng B-X, Shu Q, Wang J-F, Wang G-R, Wang D-Z, Han M-H. Biodiesel pro-duction from waste oil feedstocks by solid acid catalysis. Process Saf Environ2008;86:441–7.

[132] Sreeprasanth PS, Srivastava R, Srinivas D, Ratnasamy P. Hydrophobic, solidacid catalysts for production of biofuels and lubricants. Appl Catal A: Gen2006;314:148–59.

[133] Zabeti M, Wan Daud WMA, Aroua MK. Activity of solid catalysts for biodieselproduction: a review. Fuel Process Technol 2009;90:770–7.

[134] Corro G, Tellez N, Ayala E, Marinez-Ayala A. Two-step biodiesel productionfrom Jatropha curcas crude oil using SiO2·HF solid catalyst for FFA esterifica-tion step. Fuel 2010;89:2815–21.

[135] Deng X, Fang Z, Liu Y-H, Chang-Liu Y. Production of biodiesel from Jatrophaoil catalyzed by nanosized solid basic catalyst. Energy 2011;36:777–84.

[136] Lu H, Liu Y, Zhou H, Yang Y, Chen M, Liang B. Production of biodiesel fromJatropha curcas L. oil. Comput Chem Eng 2009;33:1091–6.

[137] Sharma YC, Singh B, Korstad J. Application of an efficient nonconventional het-erogeneous catalyst for biodiesel synthesis from Pongamia pinnata oil. EnergyFuels 2010;24:3223–31.

[138] Singh B, Bux F, Sharma YC. Comparison of homogeneous and heterogeneouscatalysis for synthesis of biodiesel from Madhuca indica oil. Chem Ind ChemEng Q 2011;17:117–24.

[139] Muthu H, SathyaSelvabala V, Varathachary TK, Selvaraj DK, Nandagopal J,Subramanian S. Synthesis of biodiesel from neem oil using sulfated zirconiavia tranesterification. Braz J Chem Eng 2010;27:601–8.

[140] Modi MK, Reddy JRC, Rao BVSK, Prasad RBN. Lipase-mediated transformationof vegetable oils into biodiesel using propan-2-ol as acyl acceptor. BiotechnolLett 2006;28:637–40.

[141] Modi MK, Reddy JRC, Rao BVSK, Prasad RBN. Lipase-mediated conversion ofvegetable oils into biodiesel using ethyl acetate as acyl acceptor. BioresourTechnol 2007;98:1260–4.

[142] Royon D, Daz M, Ellenrieder G, Locatelli S. Enzymatic production ofbiodiesel from cottonseed oil using t-butanol as a solvent. Bioresour Technol2007;98:648–53.

[143] Su E, Du L, Gong X, Wang P. Lipase-catalyzed irreversible transesterificationof Jatropha curcas L. seed oil to fatty acid esters: an optimization study. J AmOil Chem Soc 2011;88:793–800.

[144] Tamalampudi S, Talukder MR, Hama S, Numat T, Kondo A, Fukuda H. Enzy-matic production of biodiesel from Jatropha oil: a comparative study of

tainab

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

I.B. Bankovic-Ilic et al. / Renewable and Sus

immobilized-whole cell and commercial lipases as a biocatalyst. Biochem EngJ 2008;39:185–9.

145] Wang Y, Zhang L. Ectoine improves yield of biodiesel catalyzed by immobi-lized lipase. J Mol Catal B: Enzym 2010;62:91–6.

146] Kumar G, Kumar D, Poonam, Johari R, Singh CP. Enzymatic transesterifica-tion of Jatropha curcas oil assisted by ultrasonication. Ultrason Sonochem2011;18:923–7.

147] Shah S, Sharma S, Gupta MN. Biodiesel preparation by lipase-catalyzed trans-esterification of Jatropha oil. Energy Fuels 2004;18:154–9.

148] Kumari V, Shah S, Gupta MN. Preparation of biodiesel by lipase-catalyzedtransesterification of high free fatty acid containing oil from Madhuca indica.Energy Fuels 2007;21:368–72.

149] Shah S, Gupta MN. Lipase catalyzed preparation of biodiesel from jatropha oilin a solvent free system. Process Biochem 2007;42:409–14.

150] Chattopadhyay S, Karemore A, Das S, Deysarkar A, Sen R. Biocatalyticproduction of biodiesel from cottonseed oil: standardization of processparameters and comparison of fuel characteristics. Appl Energy 2011;88:1251–6.

151] Li Z, Li X, Wang Y, Wang Y, Wang F, Jiang J. Expression and characterizationof recombinant Rhizopus oryzae lipase for enzymatic biodiesel production.Bioresour Technol 2011;102:9810–3.

152] Su E, You P, Wei D. In situ lipase-catalyzed reactive extraction of oilseeds withshort-chained dialkyl carbonates for biodiesel production. Bioresour Technol2009;100:5813–7.

153] Zheng L, Li D, Jike L, Xiaolei G, Zixin Y, Tianwei T. Enzymatic synthesis offatty acid methyl esters from crude rice bran oil with immobilized Candidasp. 99-125. Chin J Chem Eng 2010;18:870–5.

154] Kumari A, Mahapatra P, Garlapati VK, Banerjee R. Enzymatictransesterification of Jatropha oil. Biotechnol Biofuels 2009;2(1),doi:10.1186/1754-6834-2-1.

155] Gog A, Roman M, Tos M, Paizs C, Irimie FD. Biodiesel production using enzy-matic transesterification – current state and perspectives. Renew Energy2012;39:10–6.

156] Tang Z, Wang L, Yang J. Transesterification of the crude Jatropha curcas L. oilcatalyzed by micro-NaOH in supercritical and subcritical methanol. Eur J LipidSci Technol 2007;109:585–90.

157] Campanelli P, Banchero M, Manna L. Synthesis of biodiesel from edible,non-edible and waste cooking oils via supercritical methyl acetate transes-terification. Fuel 2010;89:3675–82.

158] Hawash S, Kamal N, Zaher F, Kenawi O, El Diwani G. Biodiesel fuel from Jat-ropha oil via non-catalytic supercritical methanol transesterification. Fuel2009;88:579–82.

159] Chen CH, Chen WH, Chang CMJ, Lai SM, Tu CH. Biodiesel productionfrom supercritical carbon dioxide extracted Jatropha oil using subcriti-cal hydrolysis and supercritical methylation. J Supercrit Fluids 2010;52:228–34.

160] Chen CH, Chen WH, Chang CMJ, Setsua I, Tu CH, Shieh CJ. Subcritical hydrolysis

and supercritical methylation of supercritical carbon dioxide extraction ofJatropha oil. Sep Purif Technol 2010;74:7–13.

161] Rathore V, Madras G. Synthesis of biodiesel from edible and non-edible oils insupercritical alcohols and enzymatic synthesis in supercritical carbon diox-ide. Fuel 2007;86:2650–9.

le Energy Reviews 16 (2012) 3621– 3647 3647

[162] Varma MN, Madras G. Synthesis of biodiesel from castor oil and linseed oil insupercritical fluids. Ind Eng Chem Res 2007;46:1–6.

[163] Valle P, Velez A, Hegel P, Mabe G, Brignole EA. Biodiesel production usingsupercritical alcohols with a non-edible vegetable oil in a batch reactor. JSupercrit Fluids 2010;54:61–70.

[164] Shuit SH, Lee KT, Kamaruddin AH, Yusup S. Reactive extraction of Jatrophacurcas L. seed for production of biodiesel: process optimization study. EnvironSci Technol 2010;44:4361–7.

[165] Stamenkovic OS, Todorovic ZB, Lazic ML, Veljkovic VB, Skala DU. Kinet-ics of sunflower oil methanolysis at low temperatures. Bioresour Technol2008;99:1131–40.

[166] Avramovic JM, Stamenkovic OS, Todorovic ZB, Lazic ML, Veljkovic VB. Empiri-cal modeling the ultrasound-assisted base-catalyzed sunflower oil methanol-ysis kinetics. Chem Ind Chem Eng Q 2012:53, doi:10.2298/CICEQ110705053A[online-first].

[167] Bikou E, Louloudi A, Papayannakos N. The effect of water on the trans-esterification kinetics of cotton seed oil with ethanol. Chem Eng Technol1999;22:70–5.

[168] Al-Zuhair S. Production of biodiesel by lipase-catalyzed transesterification ofvegetable oils: a kinetics study. Biotechnol Progr 2005;21:1442–8.

[169] Dossat V, Combes D, Marty A. Lipase-catalyzed transesterification of high oleicsunflower oil. Enzyme Microb Technol 2002;30:90–4.

[170] Pessoa FLP, Magalhães SP, de Carvalho Falcão PW. Production of biodieselvia enzymatic ethanolysis of the sunflower and soybean oils: modeling. ApplBiochem Biotechnol 2010;161:238–44.

[171] Xu Y, Du W, Liu D. Study on the kinetics of enzymatic interesterification oftriglycerides for biodiesel production with methyl acetate as the acyl accep-tor. J Mol Catal B: Enzym 2005;32:241–5.

[172] Torres CF, Moeljadi M, Hill Jr CG. Lipase-catalyzed ethanolysis of fish oils:multi-response kinetics. Biotechnol Bioeng 2003;83:274–81.

[173] Torres CF, Toré AM, Fornari T, Senoráns FJ, Reglero G. Ethanolysis of a wastematerial from olive oil distillation catalyzed by three different commerciallipases: a kinetic study. Biochem Eng J 2007;34:165–71.

[174] Lim S, Hoong SS, Teong LK, Bhatia S. Supercritical fluid reactive extraction ofJatropha curcas L. seeds with methanol: a novel biodiesel production method.Bioresour Technol 2010;101:7169–72.

[175] Shuit SH, Lee KT, Kamaruddin AH, Yusup S. Reactive extraction and in situesterification of Jatropha curcas L. seeds for the production of biodiesel. Fuel2010;89:527–30.

[176] Su EZ, Zhang MJ, Zhang JG, Gao JF, Wei DZ. Lipase-catalyzed irreversible trans-esterification of vegetable oils for fatty acid methyl esters production withdimethyl carbonate as the acyl acceptor. Biochem Eng J 2007;36:167–73.

[177] Yustianingsih L, Zullaikah S, Ju Y-H. Ultrasound assisted in situ production ofbiodiesel from rice bran. J Energy Inst 2009;82:133–7.

[178] Refaat AA, El Sheltawy ST, Sadek KU. Optimum reaction time, performanceand exhaust emissions of biodiesel produced by microwave irradiation. Int JEnviron Sci Technol 2008;5:315–22.

[179] Soriano Jr NU, Venditti R, Argyropoulos DS. Biodiesel synthesis via homoge-neous Lewis acid-catalyzed transesterification. Fuel 2009;88:560–5.

[180] Ngaosuwan K, Mo X, Goodwin Jr JG, Praserthdam P. Effect of solvent on hydrol-ysis and transesterification reactions on tungstated zirconia. Appl Catal A: Gen2010;380:81–6.